WEDGE

USER MANUAL
Copyright 2021 © Trimble Inc.

All rights reserved. The information contained herein may be changed without prior notice. Names and
data used in examples herein are fictitious unless otherwise noted. No part of this document may be
reproduced or transmitted in any form or by any means, electronic or mechanical, for any purpose,
without the express written permission of Trimble Inc. While every care has been exercised in
compiling the data contained in this publication, Trimble Inc. accepts no liability for any errors in this
publication or for damages arising from the use of the information included in this publication. The
Wedge (Wedge) trademark used in this publication is a trademark of Trimble Inc. All other trademarks
are the property of their respective holders. Trimble Inc. disclaims any proprietary interest in the
trademarks and names of others.

Trimble Inc.

[Link]

[Link]
Table of Contents
1. Introduction .............................................................................................................................. 1
2. User information ....................................................................................................................... 2
Software users ............................................................................................................... 2
Typography .................................................................................................................... 2
3. Software components .............................................................................................................. 3
Opening window ............................................................................................................ 3
Menu bar ........................................................................................................................ 3
Shortcut bar ................................................................................................................... 5
4. Viewing diagrams ..................................................................................................................... 7
Opening diagrams.......................................................................................................... 7
Canvas ........................................................................................................................... 7
5. Viewing trends ........................................................................................................................ 11
Opening Trend windows with a single trend ................................................................ 12
Opening Trend windows with multiple measurements ................................................ 13
Limiting and screening trends ...................................................................................... 16
Deleting time intervals ................................................................................................. 16
6. Editing diagrams .................................................................................................................... 19
Using drawing tools...................................................................................................... 19
Adding measurements to a diagram ............................................................................ 20
Measurement properties .............................................................................................. 21
Adding and editing components .................................................................................. 23
Adding background images ......................................................................................... 30
Copying components from a diagram to another ........................................................ 31
Removing measurements ............................................................................................ 31
Creating or deleting measurement groups .................................................................. 32
Measurement group properties .................................................................................... 32
7. Session-specific settings ........................................................................................................ 35
Selecting measurement groups ................................................................................... 35
Selecting time intervals ................................................................................................ 35
Start/exit on-line display ............................................................................................... 37
Selecting time levels .................................................................................................... 37
Selecting grades .......................................................................................................... 38
8. Special trend functions ........................................................................................................... 41
Adding and removing user annotations ....................................................................... 41
Trend graduations ........................................................................................................ 42
 Using background colour to study measurements ...................................................... 45
Trend window settings ................................................................................................. 47
9. Calculated measurements ..................................................................................................... 49
Use of calculated measurements ................................................................................ 49
Collection of Formulas dialog ...................................................................................... 50
10. Reference time ....................................................................................................................... 55
Choosing reference times ............................................................................................ 55
Removing reference times ........................................................................................... 56
11. Compensation of delays......................................................................................................... 57
Delay model ................................................................................................................. 57
Building delay models .................................................................................................. 58
Delaying Trend plots .................................................................................................... 64
Adding and removing delayed user annotations ......................................................... 65
12. Data analysis .......................................................................................................................... 67
Uses of analysis tools .................................................................................................. 67
Data preprocessing methods in Wedge....................................................................... 71
Analysis tools ............................................................................................................... 74
Statistical process control (SPC) ................................................................................. 75
Statistics ....................................................................................................................... 77
XY plot ......................................................................................................................... 78
Histogram ..................................................................................................................... 79
Correlation matrix......................................................................................................... 81
Best correlations .......................................................................................................... 82
Cross-correlation.......................................................................................................... 83
Periodic fluctuations ..................................................................................................... 85
Searching for fluctuation source (MAR) ....................................................................... 87
Non-recurring phenomena ........................................................................................... 95
Waveform matching ..................................................................................................... 95
13. Principal Component Analysis (PCA) .................................................................................... 98
Theory of PCA modelling ............................................................................................. 98
Creating PCA models .................................................................................................. 99
Using PCA models ..................................................................................................... 100
Deleting PCA models ................................................................................................. 107
14. Macros.................................................................................................................................. 108
Recording macros ...................................................................................................... 108
Executing macros ...................................................................................................... 109
Saving macros ........................................................................................................... 109
 Editing macros ........................................................................................................... 109
Deleting macros ......................................................................................................... 110
Background macros ................................................................................................... 110
15. Printing ................................................................................................................................. 112
Page composition planning ........................................................................................ 112
Adding basic report templates ................................................................................... 113
Adding table report templates .................................................................................... 114
Adding printables ....................................................................................................... 116
16. Reports ................................................................................................................................. 118
Using the report tool................................................................................................... 118
Scaling of trend plot Y-axes in reports ....................................................................... 124
17. Saving measurement data ................................................................................................... 126
Saving measurement data ......................................................................................... 126
Copying measurement data onto the clipboard ......................................................... 127
18. Importing data from a worksheet ......................................................................................... 128
Data import ................................................................................................................ 128
Management of imported data sources ..................................................................... 131
19. User-specific settings ........................................................................................................... 133
Process Diagrams...................................................................................................... 133
Time intervals ............................................................................................................. 135
Time Settings ............................................................................................................. 138
Main Window Settings ............................................................................................... 140
Servers ....................................................................................................................... 142
1. Introduction
Wedge is a software tool designed for the management of process data and the
analysis of process fluctuations. The software is typically used to reduce quality
fluctuations in the final product, as well as to increase production efficiency.
The mathematical tools provided by Wedge enable production engineers and
operators to systematically monitor and analyse process changes. In method
development we have combined the expertise of process specialists and
mathematicians. Mathematical analyses are packaged in powerful tools for
resolving process-related issues. These tools include, among others, MAR analysis
(Multichannel AutoRegressive modelling), which facilitates easier determination of
the causes of process fluctuations, reporting and waveform matching.

With Wedge you can:

• Condense large amounts of process data into useful information that supports
decision-making
• Combine measurements and view them from different perspectives with plots
• Rapidly identify disturbances and determine their causes.

The effective use of Wedge:

• Enhances your understanding of the process
• Facilitates easier detection of process disturbances
• Facilitates speedy resolution of process-related issues
• Shortens the effective time of disturbances
• Ensures uniform quality of the product
• Lengthens the life-span of your machinery through rapid detection of equipment
faults
• Increases productivity.

All Wedge software tools have been adapted to the special characteristics of each
type of process data: data preprocessing and grade dependency can be enabled
for analysis methods. Missing data points do not disturb the analysis.

1
2. User information
This is the User Manual for Wedge. This User Manual deals with the basic use of
Wedge.

Software users
The primary users of the software are, for example, operators and process
engineers. An administrator responsible for software maintenance functions must
also be assigned. The persons mentioned above shall hereafter be referred to as
the user and the administrator.

Typography
The following typographic conventions have been used in this guide:

Typography Example

The names of user Edit menu

interface menus, dialogs,
commands and buttons
are shown in bold.

Any references to chapters For instructions, see the

or subchapters are shown chapter Printing.
in italics.

Any strings entered by the Enter a string (e.g.,

user in a text box are "tmp1").
shown in quotation marks.

2
3. Software components
This chapter describes the opening window and the main window of the software.

Opening window
When you start Wedge, the front page is displayed.

Figure 3.1 Opening window

Under the Title bar of the front page, you will see a Menu bar containing the
commands that control software functions. Under the Menu bar, you will see a
Shortcut bar containing list boxes that make using the software easier.

Figure 3.2 Title bar, Menu bar and Shortcut bar

Menu bar
The figures in this chapter show the main menus of Wedge.

3
File menu
The commands in the File menu allow you to open a diagram, save a diagram,
print selected information and exit the software.

Figure 3.3. File menu

Edit menu
The commands in the Edit menu allow you to edit diagrams, measurements and
settings.

Figure 3.4. Edit menu

Options menu
The commands in the Options menu allow you to edit the screening settings of
measurement data and the SPC limits of trends.

Figure 3.5. Options menu

Tools menu
The commands in the Tools menu allow you to create and run macros, select the
reporting interval, use the collection of Matlab formulas, save measurement data as
a text file, read data from Excel and set the monitoring settings.

4
 Figure 3.6. Tools menu

Analyses menu
The commands in the Analyses menu allow you to use analysis tools and view, for
example, measurement statistics, XY plots and cross-correlations, as well as
search for the source of periodic fluctuations.

Figure 3.7. Analyses menu

Help menu
The About Wedge command in the Help menu opens the About dialog. The
About dialog box displays the copyright information and current version information
of the software.

Figure 3.8. Help menu

Shortcut bar
The figures in this chapter show the Shortcut bar of Wedge.

5
Figure 3.9. Shortcut bar

Under the Menu bar, you will see the Shortcut bar. The Shortcut bar contains list
boxes for easier use of the following functions:
• Selecting measurement groups
• Setting time intervals
• Selecting time levels
• Selecting the process state/grade dependency
• Back to front page
• Stop a function
• Select the Edit mode

For more information on the list boxes in the Shortcut bar, see Chapter 7. Session-
specific settings.

When you open a diagram, the software opens the main window, which displays
the canvas and the measurements list menu.

6
4. Viewing diagrams
This chapter deals with opening and viewing diagrams. A process diagram is a
model of a real process consisting of image and measurement components.

Opening diagrams
You can open a diagram by clicking one of the links on the front page or from the
File menu.

Use the Open command to open a diagram that has not been set as a link on the
front page. The file format for files opened in Wedge is Wedge Process Diagram
(*.wproc or *.wprx). This format cannot be changed. Diagrams can be opened via
the Wedge server or from your own local computer.
1. Choose Open in the File menu. The Open dialog is displayed.

Figure 4.1. Opening the Process Diagram window

2. Select a directory from the Look in list box. The directory can be on the Wedge
server, in the My Diagrams directory (this directory is also on the server) or in the
My computer directory (local computer).
3. Click the name of a folder or a file in the File list.

4. Click Open to open the selected diagram. Click Cancel to cancel the selection.

Canvas
In the main window, you can see the canvas and the measurements list menu (on
the left-hand side of the window) in addition to the Menu and Shortcut bars. Be
careful not to close the main window completely during use as Wedge interprets
closing the window as quitting the application. The following sections briefly

7
 describe the main parts of the main window: the canvas and the measurements
list.

Figure 4.2. Main window

Depending on the selection in the Select a Measurement Group list box, the main
window displays either the entire process diagram (a diagram) or the group created
under it. A diagram is a model of a real process consisting of image and
measurement components. The measurement components (measurements) are
used to display data as trends. A data indicator gives a graphical representation
of the end point of the selected time interval. The image component contains no
data; rather, its function is to help the user to follow the process.

Viewing diagrams
You can open only one diagram at a time. If the diagram is too large to be
displayed entirely in the window, you can use the horizontal and vertical scroll bars
to view it. The size of the diagram can also be reduced by using the mouse scroll
wheel while pressing the Ctrl key or by pressing the Ctrl and +/- keys
simultaneously.

8
Finding measurements in a diagram
The Find measurements function helps you to find measurements. The quick
search function starts automatically when you press a letter on your keyboard. The
quick search box will appear in the upper right corner of the main window.

Figure 4.3. Quick search box

The quick search box shows the number of measurements found with the search
criteria in a diagram and which measurement is selected at the given moment.
Press Enter to open the Quick trend window. Use the Previous/Next buttons to
change the selected measurement. To close the quick search box, press the Close
button.

Measurements list
On the left-hand side of the main window is a menu with the options
Measurements in diagram, All measurements in server and Hidden.

9
Figure 4.4. Measurements list

The Measurements in diagram menu displays the names of the measurement

components in the current diagram or measurement group. You can expand or
shrink the measurements list view by dragging the split bar between the
measurements list and the canvas. You can change the order of the
measurements list contents by clicking the headers of each column.

The appearance of the measurements displayed in the measurements list has the
following meaning:
• Normal text: the measurement configuration is in order and data can be
retrieved for processing.
• Red text: there is an error in the measurement settings and no data can be
retrieved for processing.
The measurements list shows the measurements for the measurement group
currently in use. You can search for measurements with the search field at the
bottom of the measurements list. In the Column selection menu, you can select
how many columns are displayed in the measurements list. The search function
searches all visible columns for the measurements.

The columns are:

• Name
• Unit
• Description
• Data source
• Measurement data source position.
The Search options menu allows you to enable or disable case sensitivity and
wild card characters in search strings.

10
5. Viewing trends
Once you have selected a diagram, you can view the measurement data. The most
popular presentation format used in process data is the trend graph. A trend graph
refers to a graphical representation of the consecutive points of a time range within
a single measurement, where the consecutive data points are connected to each
other by a line. Trends enable you to view process events, such as process
component temperatures, pressures or chest levels.

Opening quick trend graphs

Click the name of a measurement to open a quick trend graph and statistics at the
bottom of the main window.

Figure 5.1. Quick trend graph in the main window

The trend time interval shortcuts are displayed at the bottom of the quick trend
graph. You can navigate the time intervals forwards or backwards in quarter or full
intervals. In the separate calendar view, you can choose any arbitrary time range to
single-day accuracy or choose a so-called favorite time interval.

11
 Select an interval in the trend graph with your mouse to display the statistics for the
selected interval as well as the time and value differences for the start and end
points of the interval.

Figure 5.2. Selected interval in a quick trend graph

Press the button at the upper right corner of the quick trend graph to display the
quick trend graph in a separate window. To close the quick trend graph, click the
canvas or the button at the right corner of the graph.

Opening Trend windows with a single trend

Double-click the desired measurement in the diagram. A separate Trend plot
window will appear. The trend time interval shortcuts are displayed at the bottom of
the window.

The following figure shows the boxes and buttons in the Trend window.

12
 Figure 5.3. Trend window functions

Opening Trend windows with multiple

measurements
To open a Trend window with multiple measurements:

1. Do one of the following to choose the desired measurements on the diagram:

• Move the mouse pointer near to the first measurement you want to select. Hold
the left mouse button down and drag a rectangle around the measurements.
Release the left mouse button.
• Hold the Ctrl key down and click the measurements to select them.
2. When the measurements are selected, do one of the following to open the trend
graphs:
• Click the Trend button that appears near to the right-hand edge of the
measurement name

Figure 5.4. Select buttons for trend graph and statistics

• Right-click the mouse on a white area of the diagram and select Separate
Trends or Combined Trends in the Trends and Spectra menu.

13
 Figure 5.5. Trends and Spectra menu

The Separate Trends view opens the Trend window, showing the trends of the
selected measurements as separate trend plots. Measurement names are
displayed above the trend plots.

The Combined Trends view opens the Trend window, showing the trends of the
selected measurements in a single trend plot in different colours. Measurement
names and the corresponding trend colours are displayed beneath the coordinates.
Should you wish to view more than six trend plots, a scroll bar will appear on the
right-hand edge of the Trend window. With the scroll bar, you can browse all of the
trend plots you have selected. If you do not select any measurements from the
diagram and then open the Separate Trends command, all measurement trend
plots are searched for viewing. They can be browsed in groups of six
measurements using the scroll bar.

Combined trends view: special functions

When you select the Combined Trends function after selecting multiple
measurements, trend graphs are displayed as groups in several axes in the trend
plot. The graphs are grouped so that measurements with near-similar fluctuation
ranges are placed on the same axis. The program displays either one axis at the
left-hand edge of the trend plot or one at both edges of the trend plot. If there are
more than two groups, these additional measurements "float" in the plot so that
their fluctuation is shown as large as possible.

14
 Figure 5.6. Combined Trends view and menus

You can move a trend graph from a scale to another by left-clicking the mouse on
the name of the trend and selecting "Attach to left axis ", " Attach to right axis " or
"Detach from both axes" in the menu.

In the same menu, you can change the trend graph colour and the thickness of the
trend line. You can also use the menu to hide a trend graph or remove it from a
plot.

Left-click at the bottom of a trend plot to open a shorter menu which allows you to
copy an image to the clipboard and from there to Word or to move all
measurements to the left scale.

Adding trend plots to an open Trend window

You can add new trend windows to an open Trend window using the drag-and-
drop mouse function. To add a new trend plot to an open Trend window:
1. Select the desired measurements from the diagram, open the Trend window
(either Separate Trends or Combined Trends command).

2. Choose one or more new measurements from the diagram. Hold down the left
mouse key and drag the mouse pointer to the desired point in the open Trend
window. A thin red horizontal line is displayed between the Trend windows.

3. Release the left mouse key; the new trend plots will open in place of the red line.
You can add new diagrams above, between and below the open trend plots

15
Limiting and screening trends
To view trends in detail, you can zoom in on a specific part of the trend plot using a
limiter. You can also use the Trend window tools for measurement value
screening.
When you left-click the trend plot, the program creates an X or Y limiter
automatically on the trend plot. The type of the limiter depends on the direction in
which the mouse pointer is drawn.
When the trend is limited in the X-axis direction, a new scroll bar is created below
the plots on the trend display. Use the scroll bar to navigate forwards or backwards
along the trend time range.

The figure below presents trend limiting and data point screening in the Trend plot.

Figure 5.7. Trend plot

Deleting time intervals

The Delete Time Interval function can be used, for example, when you want to
remove:
• A time interval that gives clearly incorrect data (for example, breaks)
• A time interval containing no measurement data of interest.

Removing single time intervals from a trend plot

To remove a single time interval from a trend plot:
• Select a time interval from the X-axis.

16
 • From the trend plot tools, select either Remove the selected time interval
from data or Remove other than the selected time interval from data.

Removing all time intervals that exceed limits

To delete all time intervals where the selected measurement exceeds certain
values:
1. From the Y-axis of the trend plot, select the values that limit useful measurement
signal values.

2. From the trend plot tools, select Delete Exceeding Values.

Figure 5.8. Delete Exceeding Values function

Restoring time intervals

To undo time interval deletions, right-click on the trend plot to open the Edit menu
of the trend plot and select Cancel data deletions.

You can also undo the deletion of time intervals by selecting a new time interval for
use.

Measurement information
To open the information of a measurement in a separate window, press the
Measurement information button.

Removing trend plots in a multiple trends window

If the Trend window contains several trend plots, you can use the button at the
right-hand edge of the trend window to close the plots one at a time.

17
 Figure 5.9. Trend plot properties

Keeping Trend windows

At the right corner of the trend window is the Keep this trend locking bolt. This
function locks the Trend window on the display, enabling other Trend windows to
be subsequently opened. The locking bolt can be used to lock any given trend
window, and the number of open and locked trend windows is unlimited. To
activate the locking function, click it with the left mouse button.

18
6. Editing diagrams
This chapter discusses creating and editing a diagram.

In this chapter, you will learn the following:

• Using drawing tools
• Adding new components into a diagram or its subprocess
• Creating new diagrams
• Copying components from one diagram or subprocess to another
• Defining measurement group-specific properties for measurement trends and
for the indexing of measurement components.

Using drawing tools

Editing changes made to a measurement group are not automatically updated in
measurement groups further up in the hierarchy. In other words, the measurement
groups are independent of each other: they can contain measurement and drawing
components which the measurement groups at the main diagram level or those
next to, below or above each other in the hierarchy do not have. The diagram can
be changed only in edit mode.

To edit a diagram, click the shortcut key at the right corner of the diagram window
to open the Edit menu. A drawing toolbar will appear on the right in the main
window, which also indicates that the edit mode is active. The toolbar contains the
drawing tool buttons described below.

19
 Figure 6.1. Drawing tools

Adding measurements to a diagram

On the left-hand side of the main window is a menu with the options
Measurements in diagram and All measurements in server.
The All measurements in server menu displays all the server-based
measurements that can be added to a diagram. You can choose to view all
measurements on a server or measurements within an individual data source.
Measurements can also be chosen according to a specific Process area if one has
been defined.

You can expand or shrink the measurements list view by dragging the split bar
between the measurements list and the canvas.

The measurements list columns are added by left-clicking the Choose columns in
list button. Select the desired columns by left-clicking the menu that opens up. If
you have chosen to view All measurements in server, you can add the columns
Name in diagram and Process area.

20
 Figure 6.2. Adding measurements list columns

To add measurements to a diagram from All measurements in server:

1. Select All server measurements for the measurements list.

Figure 6.3. Search fields

2. Enter a string (for example "ref" or "pH_") into the entry field. The search item
can simultaneously be tag, unit or description.

3. The measurements matching the search terms are shown on the measurements
list.

4. Drag a measurement to the desired point in the diagram.

Measurement properties
To change the settings for the measurement component shown in the diagram
dialog, right-click on the component to open a menu and select Component
properties….

21
Figure 6.4. Component properties, General tab

1. To change the measurement name in the diagram, enter a new name into the
Name: field.

2. You can change the Data source of the measurement and with the drop-down
menu and the Tag of the measurement by entering a new position name in the text
box. To check whether there is a tag in the already open diagram, click Browse…
If there is a tag in the opened process diagram, its label is shown in the Name in
diagram column.
3. To choose whether an icon is shown or hidden, select or deselect the Show
icon check box.

4. You can change the Measurement icon by clicking the Change... option under
Component change. The available options are:
• Basic icons: all basic software measurement icons
• All icons: basic icons component library, expanded with the Process parts icon
library
• Custom Icon
5. Use Show measurement name, Show measurement value and Show
measurement unit to set the properties for the label.

The Show measurement value option shows the last value of the measurement in
the diagram dialog. The value is updated whenever the selected time interval is
changed.

22
 6. Use as state measurement - selection imports the measurement to be used in
the Grade selection menu.

Figure 6.5. Measurement properties, Trend view tab

In the Trend display tab, you can set the minimum and maximum Y-axis scaling of
the trend for use in automatic reports. To add numeric values to the Y-axis of the
trend, select the Use maximum Y-axis resolution check box. You can also choose
the background colour for a trend plot – select a measurement from the
Background colour from measurement drop-down menu.

Adding and editing components

To add a new component to a diagram or subprocess:

1. From the Edit menu, select the Add components… command. The
Components dialog will appear.
2. Click a tab in the Components window and double-click a component to display
it in the canvas.

When a diagram is in the edit mode, you can also add components by right-clicking
the diagram and selecting New component.

23
 Figure 6.6. Add new component menu

With this menu, you can add a new Calculated measurement, New text field, New
component icon or New link to a measurement group.

To edit the component in use, right-click the component and select Component
properties from the menu.

Adding text field components to a diagram canvas

1. Double-click the text field component in the Generic tab of the Components
dialog. The Text Field dialog will appear.

Figure 6.7. Text Field dialog

2. Enter a text in the text box.

3. Set the font type and size.

4. If you want to add frames to the text field, select the Framed check box.

5. Select Add. The Text Field dialog will close and the created text field component
will appear on the canvas.

24
Adding macrolink components to a diagram canvas
1. Double-click the macrolink component in the Generic tab of the Components
dialog. The Macrolink dialog will appear.

Figure 6.8. Macrolink dialog

2. Enter a name for the macrolink in the Link name text field. This name will be
shown in the diagram.

3. Set the font type and size.

4. If you want to add frames to the text field, select the Framed check box.

5. Select a macro from the Link to macro drop-down menu.

6. Select Add. The Macrolink dialog will close and the component to run the
created macro will appear on the canvas.

Adding Linear estimation components to a diagram canvas

Linear estimation is a method used to create a linear model between explanatory
measurements and the response variable. To add a linear estimation component to
a diagram or subprocess:

1. Select the measurement component for the response variable in the diagram
and then select 1 to 5 explanatory measurement components.

2. Double-click the Linear estimation component in the Generic tab of the

Components dialog. The Linear estimation dialog will appear.

25
 Figure 6.9. Linear estimation dialog

3. Choose the Formula tab. The primary measurement component is shown in the
Estimated measurement drop-down menu. The estimated measurement y is
modelled using a linear equation in which the variable e is the modelling error.

y = b + x1*a1 + x2*a2 + x3*a3 + x4*a4 + x5*a5 + e

4. Other measurement components are shown in the Measurement fields.

5. Click the Fit coefficients button to calculate the model coefficients.

6. Click OK. The Measurement properties dialog will close and the selected linear
estimation component will be displayed on the canvas.

Adding Link to measurement group components to a diagram

canvas
To add a quick link to another subprocess of the same diagram:
1. Double-click the Link to measurement group component in the Generic tab of
the Components dialog.

26
 Figure 6.10. Link to measurement group dialog

2. From the Link target drop-down menu, select the measurement group of the
process diagram you want to link to. Close the Link target list box by clicking a
blank area on the screen or by pressing the Esc key.

3. Enter the Link label in the textbox:

4. Click OK. The Link properties dialog will close and the new link to the
measurement group will be displayed in the diagram canvas.

Adding browser pages to a diagram canvas

You can embed browser pages into a diagram. You can use browser pages to add
process-related background information into a diagram dialog, such as a report
page of a laboratory system with a browser user interface. To add a browser page
into a diagram:

1. Double-click the Web page component in the Generic tab of the Components
dialog.

Figure 6.11. Web page configuration window

2. Enter the web page address and the page size of your choice.
3. Move the page to the desired point in the diagram.

27
Adding data indicators
You can add components indicating measurement values into a diagram. These
components observe the measurement value at the end of the selected interval
and indicate this value with colors or bar length.
The image below presents the data indicators available for use.

Figure 6.12. Data indicator components

Choices:
• Single-colour column
• Two-colour column; the component has one limit value which triggers a colour
change
• Three-colour column; the component has two limit values which trigger a colour
change
• Round colour indicator with sliding colour from green to red
• Round colour indicator; the component has one limit value which triggers a
colour change
• Round colour indicator; the component has two limit values which trigger a
colour change
• Single-colour dial
• Dial, with sliding colour with three different colors
• The above choices are also available for use with calculated measurements.

To add a data indicator:

1. Double-click a component in the Data indicators tab.

2. Move the component to the desired point.

3. Right-click the component and select component properties.

4. Set the parameters of your choice for the indicator.

28
 Figure 6.13. Parameters for two-color columns

You can also change an existing measurement component in the diagram into a
data indicator component by changing the icon of the component in the component
properties.

Adding analysis components

You can add analysis result windows into a diagram. These results are
automatically updated when the interval changes and you do not need to open the
results separately to view them.

To add a Statistics window of selected measurements into a diagram:

1. Select the measurements of your choice.

2. From the Edit menu, select the Add components… command.

3. Click the Analysis components tab in the Components window.

4. Double-click the Statistics component.

5. Move the component to the desired point.

Figure 6.14. Statistics information view embedded into a diagram

Histograms are another ready-made example. You can also add a general
Analysis component, which allows you to create new embedded views. For more
information, contact the software supplier.

Adding passive image components

You can add passive image components to a diagram using the Components
dialog tabs. Passive components cannot be connected to a data source.

29
 To add a passive image component:

1. From the Edit menu, select the Add components… command.

2. Select a tab from the Components dialog.

3. Double-click the desired component.

4. Move the component to the desired point.

Adding background images

An image can be inserted into the background of process diagrams, onto which
measurement components and passive image components can be placed.
To add a new background image into a diagram:

1. Copy the background image onto the clipboard of the Windows operating
system. Right-click, and from the menu that opens up, select Edit mode of
background images.

2. Import the background image via the Paste background image command.
3. If you are using several different background images, they must all be inserted
into each measurement group of the diagram separately. In other words, the
process diagram structure for background images can be as follows: general image
of the production line at the main diagram level, images of individual process parts
in the measurement subgroups of the diagram.

Figure 6.15. Paste background image command

4. A single measurement group can have several different background images. If

you add more than one background image, you can change the image overlap
shown in the diagram background.

30
 Figure 6.16. Background images editable command

5. To exit Background Images Editable, click the left mouse button. The
checkmark will disappear from the line.
Once the background images of the diagram have been inserted, you can begin
placing measurement components and passive image components onto them.

Copying components from a diagram to another

Process diagram components can be copied from the canvas to another when two
Wedge programs are running.

To copy diagram measurements or passive image components:

1. Set the diagram to edit mode. Select the components you wish to copy.

2. Right-click on a white area of the canvas. From the menu that opens up, select
Copy measurements. If you are copying calculated measurements, also select all
of their input measurements.
3. Open another diagram. Make sure that the program is in edit mode.

4. Right-click on the white area of the diagram and choose Paste measurements.

Removing measurements
Measurements are removed using the Remove Components drawing tool. To
delete a measurement component:

1. Select the components that you want to delete.

2. Click the Remove Components from a Group drawing tool in the drawing
toolbar. The program removes a component from a measurement group and its
subgroups.
3. You can also remove the selected measurement components by pressing the
Delete key on the keyboard. The selected components are deleted from the
diagram immediately.

31
 You can reverse the component deletion via the Undo: delete components
command in the Edit menu.

Creating or deleting measurement groups

The Create measurement group function is especially useful when you want to
limit a measurement group for viewing purposes and to reduce the number of
measurement components viewed at one time.

To create a new subprocess:

1. Select the components from a diagram.

2. Select Create. The New subprocess dialog will appear.

3. Enter the name for the new measurement group into the text box.

4. Click OK.
Use the Delete measurement group function to remove any unnecessary groups
from a diagram.

Measurement group properties

In the Group properties... – dialog, you can set properties for the measurement
groups in a diagram.

Figure 6.17. Group properties dialog

In the Group properties dialog, you can specify measurement group-specific

settings as to which graduations of trends are displayed:

32
 • Target value
• Alarms
• Lower and upper limits
• Average.

To use Group-specific settings:

1. Go to the desired measurement group.

2. Activate the Group-properties option in the measurement group that you are
editing. Select the graduations you wish to display in trends.

3. Click OK to accept the settings.

4. Save the diagram. When you reopen the diagram, the graduations you have
defined will be automatically applied to the measurement groups for which you
have defined group-specific settings.

If you do not want to use the Group-specific settings option, the measurement
graduations must be separately activated whenever the diagram is opened in
Wedge.

If you want to add a common, automatic background colour to all measurements

belonging to a given measurement group, select the option Background coloring
measurement. The background colour is saved as a diagram measurement group
property in a similar way to measurement graduations.

Measurement sorting in a measurement group

A group-specific measurement indexing order can be defined in the diagram. The
indexing order is taken into account, for example, when you use the Compare
trends with this measurement feature of the Trend window. The indexing order
also affects the order in which measurements are selected for automatic reporting
by Wedge.

To enable indexing:

1. Go to the measurement group for which you wish to change the measurement
indexing order.

2. Right-click to display the Group properties… command. Choose the

Measurement sorting tab.

3. Choose either of the following alternatives:

- Alphabetical sorting: measurements are indexed in alphabetical order.

- User-defined sorting: measurements are indexed in the order defined by the

user.

4. If you choose User-defined sorting, you can change the indexing order using the
Up and Down buttons.
5. Click OK to accept the settings.

6. Save the diagram. When you reopen the diagram, the indexing you have defined
will be automatically applied to the measurement groups for which you have
defined group-specific settings.

33
Figure 6.18. Alphabetical sorting of measurements

34
7. Session-specific settings
This chapter deals with the use of the list boxes of the Shortcut bar. The Shortcut
bar contains session-specific settings, i.e., user-specified settings that are not
permanently stored in the memory.

Figure 7.1. Shortcut bar

In this chapter, you will learn how to select the following:

• Measurement group
• Time interval
• Time level
• Certain grades from an interval.

Selecting measurement groups

A diagram must be open before selecting a measurement group. The name of the
selected measurement group is displayed in the Select a Measurement Group list
box on the left-hand side of the Shortcut bar.

A measurement group is a body of measurements that facilitates the viewing of

diagram components. Use measurement groups, for example, when you want to:
• Analyse all pH values or streams
• Analyse measurements associated with a specific subprocess
• Limit the target area of the analysis.

Measurement groups are organised in a tree-like structure. Measurement group

subprocess can be independent of the uppermost level of the diagram. This means
that the measurement group can contain components which are not found on the
upper level. Furthermore, the location of measurement group components can also
differ from the upper level.

Selecting time intervals

The time interval refers to the start and end times of the data to be processed. A
time interval, displayed in the Shortcut bar, is always selected in Wedge. The
default time interval is the Last Day, unless the function has been customised by
the user.
You can select the time interval either from the trend plot shortcuts or by clicking

35
 the Set Time Intervals list box on the Shortcut bar.

Time interval shortcuts

To select a time interval shortcut or one of the most recently used time intervals in
the Set Time Intervals list box:
• On the Shortcut bar, click the Set Time Intervals list box or the drop-down
arrow on the right-hand edge. A drop-down list will appear.
• Click to select a shortcut or one of the most recently used time intervals. The
selected time interval is displayed in the Set Time Intervals list box.

Setting custom time intervals

To define a custom time interval:
1. Click the Set Time Intervals list box and select the Set Time Intervals
dialog.

Figure 7.2. Set Time Intervals dialog

2. Select the time interval presentation format in the Time Interval Format list box.

3. Select the time interval. You can enter the numbers in the combo boxes from the
keyboard or by using the up and down arrow keys.

4. Click Apply or OK.

- Clicking the Apply button accepts the selection, but does not close the dialog.
You may continue making selections in the dialog.

- Clicking the OK button will accept the selection and close the dialog.

Time Interval Formats

The Time Interval Formats list box in the Set Time Intervals dialog shows the
following three options:
• Time interval format in dates
• Time interval format in relative times
• Time interval format in data points.
Once you have selected a time interval format, the software automatically adjusts

36
 the time interval boxes to correspond to the selected time interval format.

Start/exit on-line display

The time interval shown by Wedge can also be dynamic, i.e., it is updated as time
progresses. This means that the opened Trend window or some other view is
updated as new process data accumulates. The update rate depends on the
selected time level. To activate the dynamic trend display, left-click the Start on-
line display command in the Set Time Intervals list box.

Figure 7.3. Start on-line display

When the dynamic trend display is active, the trend display mode symbol changes
in the lower right-hand corner of the display. The dynamic trend display can also be
activated via the symbol in the bottom corner of the diagram.
The on-line display can be exited via the lower right-hand corner of the diagram
display by left-clicking the Stop on-line display button. The on-line display can
also be stopped via the Set Time Intervals list box by left-clicking the Set on-line
display button.

Figure 7.4. Stop on-line display

Selecting time levels

A time level refers to the time range between two consecutive points in the
measurement data time series, i.e., the sampling interval. When you select the time
level, you also set the sampling interval.
To change the time level format:

37
 Click the drop-down arrow of the Time Level list box to display the alternative time
levels. Select a time level.

Figure 7.5. Time Level list box

One time level, displayed in the Level list box of the Shortcut bar, is always
selected in Wedge. If you do not select a time level from the list box, Wedge will
use the user-specific default time level.

Selecting grades
All different product grades manufactured in the process are shown in the tree
structure of the Select Grade list box. They are organised according to customer
requirements and product grade. Process-specific grade codes are created by the
Wedge supplier and cannot be modified by the user. For more information on
process grade codes, contact your software administrator.

By default, Wedge uses the All data option in the Select Data list box.
You can open the lower level of the structure by clicking the plus buttons. Similarly,
you can close the lower levels of the structure by clicking the minus buttons.

Figure 7.6. Select Grade list box

Figure above shows a grouping for a paper mill. The paper grades manufactured at
the mill are grouped into three categories:
• By grade, e.g., Grade A/x 230, Grade A/x 250, Grade B/x 230 and Grade B/x
250
• By product group, e.g., A and B
• By basis weight, e.g., 230 g/m2, 250 g/m2 and 270 g/m2.

Only the product groups or grades manufactured during the selected time interval
are shown.

38
Grade dependency criteria
Grade dependency is needed when using analysis methods. Grade dependency
means that the analysis is only applied to a specific grade or product group.

To be able to use grade dependency-based functions, the diagram must include a

measurement signal that indicates the product currently produced.

To view the trend for an individual grade, select the name of the grade from the
Select Grade list box. Only the data points belonging to the grades in question are
drawn on the trend display. You can also choose a specific product group.

You can also select several different grades or groupings for simultaneous viewing.
Select multiple grades as follows:
1. Click to select a grade in the Select Grade list box.

2. To then select other grades, hold the Ctrl button down on the keyboard and click
the grades with the mouse.

Using Grade Dependency

When viewing the trend, you will see the current grade code in the upper left-hand
corner of the trend window.

Figure 7.7. Trend window with several grades selected

If the trend plot contains two or more grades for which the production time is not
consecutive, a grey vertical line is shown on the trend plot at the grade change
point on the time axis.

In addition to trends, you can enable grade dependency in the following Wedge
analyses:
• Statistics
• XY plot
• Histogram
• Cross-correlation

39
• Correlation matrix
• Best correlations
• Periodic fluctuations
• PCA.

40
8. Special trend functions
This chapter deals with the following subjects:
• Adding and removing user annotations
• Graduations dialog
• Setting graduations
• Editing Trend window properties.

Adding and removing user annotations

When you have selected with the X limiter a time interval in the trend plot that
contains interesting measurement data, you can then change the background
colour of the trend plot at the time point in question. The same time interval is
marked with the background colour in both the original Trend window and the
Trend windows that are opened later.

Enable trend background colors in a user annotation as follows:

1. Select the desired time interval with the X limiter from the trend.

2. Right-click on the trend plot and select Add user annotation.

To remove a user annotation:

1. Right-click the trend plot. Select Delete user annotations from the menu that
opens up. All background colour markings will be removed.

Figure 8.1. Trend plot background colour

41
Trend graduations
Graduations are session-specific lines that appear in the trend plot and give more
information on the trend. Wedge graduations give information on, for example,
saved limits, data cutoff points and the average. When measurement graduations
have been selected, they will remain defaults until the selection is removed.
To edit trend graduations, right-click on the trend plot and select Graduations.

Figure 8.2. Graduations dialog

Use the Graduations dialog to define the following settings:

• Data cutoff points: When this option is selected, a grey vertical line is
displayed in the trend plot at the point where data has been cut out. The
program replaces the whole cut section with a single grey line.
• The Set Limits… button: Open the Measurement Limits dialog, where you
can set measurement limits. You can also open the dialog by selecting the
Measurement Limits... command from the Settings menu.
• Saved limits: Enable the limits that you have set and saved.
o Target Value: the target value of the measurement, shown as a
green horizontal line in the trend plot.
o Lower and Upper Limits: the lower and upper limits of the
measurement values, shown as red horizontal lines in the trend
plot.
o Alarms: time intervals where the measurement values exceed the
upper or lower limits, shown with a yellow background in the trend
plot.
• Calculate from current data – average: calculates an average of the current
data and shows it as a blue horizontal line in the trend plot.

42
 • Background colour from measurement: a measurement, the trend content of
which is displayed as gradually changing background coloring in all trends. The
lightness of the blue background reflects the changes in the measurement
value: the lower the measurement value, the darker the background colour.
When a background colour is used in a trend, the variation of a selected
measurement and other trends can be compared simultaneously.

The following table contains the descriptions of all colored lines that may appear in
a trend plot.

Line Description

Grey vertical line Shows the points where data has been cut out

Red horizontal line Upper or lower limit of measurement values

Green horizontal line Measurement target value

Yellow vertical area/line Alarm: a time interval that contains values over
the upper limit or under the lower limit

Blue horizontal line Current data average

Setting measurement limits

In the Measurement Limits and Limits dialogs, you can set the limits for
measurements.

Figure 8.3. Measurement Limits dialog

The measurements selected for the setting of limits and their state dependency are
shown in the columns of the Measurement Limits dialog. State dependency
indicates whether a measurement has a single set of limits or product-specific
limits.
When a measurement name is selected from the list, a state dependency option
can be selected from the State Dependency list box.

Click the plus button in front of the measurement name to display measurement
values according to the selected State Dependency option (for example, by grade

43
 or from all data). When you click an option in the structure, the program shows the
limits for that measurement if the values have already been set:

When you click a measurement name or a grade or product group name in the
structure, the lower limit, target value and upper limit of the measurement are also
copied to the text boxes below the list. Enter values in the text boxes, if necessary.
When defining settings manually, the upper limit, lower limit (one-sided limits) or
both can be set.

Click the Change button to copy the change into the columns in the upper parts of
the dialog as well. Measurements are not updated with the new data until the data
has been approved by clicking the OK button.

Limits dialog
Measurement limits can also be calculated on the basis of the data in use. To do
this, click the Calculate button in the Measurement Limits dialog.

Figure 8.4. Limits dialog

Use the Limits dialog to select the principle that Wedge uses to calculate a lower
limit, target value and upper limit for selected measurements.

If the measurement data can be considered normally distributed, the Normal

distribution option can be used to calculate the limits.

The Normal distribution option calculates the average and standard deviation and
defines the values shown in the table. The user enters a K value. (The default K
value in Wedge is 3. The value is normally selected between 1 and 3.) The value is
a factor that the program uses to indicate the standard deviation of the data.

Values Calculation principle

Target value Measurement data average

Lower limit Target value – K * data standard deviation

Upper limit Target value + K * data standard deviation

You can choose the K value and enter it in the Measurement limits calculation
dialog. The statistical interpretation of the K value is as follows:
• 99.7% of data points lie within 3 standard deviations of the average.
• 95.4% of data points lie within 2 standard deviations of the average.

44
 • 68.3% of data points lie within 1 standard deviation of the average.

The above interpretation is only valid for normally distributed data. When the upper
and lower limits are calculated on the basis of a normal distribution, they are
located symmetrically on both sides of the average.

The values are calculated from the measurement data. If the data is not normally
distributed, the limits can be calculated according to the Measurement data
distribution. The user enters a Y percentage value in the text box. The value
indicates the share of active data remaining within limits. Thus, the share of data
remaining outside limits is as follows:

X% = 100% - Y%

The Measurement data distribution option sorts the data points by size. The
upper and lower limits are determined as specified in the following table:

Values Calculation principle

Target value Median (middle data point)

Lower limit X/2% of the lowest values remain under the

lower limit.

Upper limit X/2% of the lowest values remain above the

lower limit.

If the limits are calculated based on the actual distribution, they are usually not
positioned symmetrically.

Using background colour to study measurements

Use background colour to show information of a measurement in any trend plot.
The background colour can give information, for example, on process disturbance
start times.

To add background colour:

1. Right-click on the trend plot and select the Graduations… command.

2. Enter the name of a measurement in the Background colour from

measurement text box. All measurements contained in the diagram can be
browsed in alphabetical order via the drop-down menu.

3. Click OK to accept the selections. The Graduations dialog will close.

45
Figure 8.5. Background colour from measurement

In the above figure, the broke percentage measurement BROKE_% has been
selected as the background trend. Once the selection is made, all trend plots show
the following values of the BROKE_% measurement signal:
• High values are shown as light colors.
• Low values are shown as dark colors.

46
Trend window settings
The Properties option in the Edit Trend Plot menu opens the following dialog:

Figure 8.6. Trend window settings

Trend plot grid

The grid consists of vertical lines displayed behind the trend plot and spectrum plot.
The lines help the user to follow the time intervals of the diagram.

To activate the grid, click the Show grid check box.

To select the number of grid lines shown behind the trend plot or spectrum plot:

1. Right-click on the trend plot and select Graduations….

2. Select one of the two radio buttons.

• The number of grid lines depends on the window size: the program adds a
suitable number of grid lines based on the window size.
• Fixed number of grid lines: enter the number of grid lines in the text box. The
program adds the selected number of grid lines in all trend plots.
3. Click OK. The Properties dialog will close. The selection remains active
throughout the session.

Trend window’s time axis format

The Trend window’s default time interval format as displayed in trend plots in the
date format. The other available time axis formats are data points or seconds.

The data point time interval format defines the time interval by displaying the
sampling points from the trend's initial data point onwards. The data points are
presented in the XXXXXXX format, with the sampling points counted from 0
9999999.

The seconds time interval format defines the time interval by displaying the location

47
 of sampling points in seconds in relation to the trend's initial data point. The trend's
first sampling point is at the 0 seconds point.

The selection remains active throughout the session.

Figure 8.7. Trend window’s time axis format

Trend plot's Y-axis in logarithmic format

The trend plot's Y-axis can be changed to logarithmic format. This presents the
trend plot's sample point values as a decimal logarithm. The logarithm can only be
derived from a positive figure, so if the trend plot contains zero or negative values,
it cannot be made logarithmic.
The decimal logarithm (lg) for figure x refers to the power to which ten must be
raised to obtain x. For example, lg(100) = 2, because 102 = 10*10 = 100. This
format can be used to compare variations between trend plots when comparing
different measured variables for which the absolute values greatly differ (e.g.,
temperature compared to pressure).

To convert the trend plot's Y-axis to logarithmic format:

1. Right-click on the trend plot and select Graduations….

2. Click the Other settings – show logarithmic values check box.

3. Click OK. The Properties dialog will close. The selection remains in use only
while the trend plot is kept open.

If the same Trend window is closed and reopened, the Y-axis will be displayed in
the normal manner. If you open the Trend window of several trend plots, the Y-axis
of each diagram must be separately changed to logarithmic.

48
9. Calculated measurements
This chapter deals with calculated measurements and presents the collection of
Matlab formulas used in Wedge.

The chapter deals with the following issues:

• Basics of calculated measurement
• Collection of Formulas dialog
• Adding calculated measurements from the dialog.

Use of calculated measurements

A calculated measurement is not a physical sensor in the measured process as
such, rather it is a measurement and analysis tool that calculates the values of its
variable on the basis of values provided by actual measurement sensors.
Calculated measurements can be used in a variety of ways. Typical applications
include unit conversions, data combinations and various balance calculations.

When a calculated measurement is used, the time series resulting from the
calculation must match in length the original time series used for the calculation.
Short and simple formulas should be entered directly in the Formula field of the
Component properties dialog of the calculated measurement. Longer, more
complex formulas should be run on disk using their own functions (m-file) and then
named in the Formula field of the Calculated measurement dialog. Instructions
for m-file use are available from the Wedge administrator or the system supplier.

Calculated measurements can be added in three different ways. You can add a
calculated measurement either via the Edit menu or in the diagram edit mode by
right-clicking and selecting New component – new calculated measurements
from the menu that opens up. The easiest way, however, is to use Wedge's
existing collection of Matlab formulas, where you can add calculated
measurements by clicking their links.

Using calculated measurements requires knowledge of Matlab syntax and

functions. The most commonly used operators are + (addition) and - (subtraction).
The calculation formula is a Matlab expression, for example:
• x1+x2
• x1+12*x2.*x3-x1./(3*x2)
• sqrt(x1)

Basic calculations are performed using the following operators:

• + (addition)
• - (subtraction)
• .* (pointwise multiplication)
• ./ (pointwise division)

49
 It is important to note that the operator.* or./ must be used in order for the
multiplication or division to be performed pointwise. For example:
• x1.*x2

Normal calculation order rules apply.

The comparison operators are:

• < (less than)
• > (greater than)
• == (equal to)
• ~= (not equal to)
• & (and)
• | (or)

The following built-in Wedge variables can be used in the calculation formula:
• len (length in data points)
• dT (time level in seconds; e.g., 1 minute at the dT time level = 60)
• T (number of seconds since the beginning of 1970)

Some general commands:

• diff (differentiation);

The difference between two consecutive data points. N.B.! The resulting time
series is 1 data point shorter than the original time series. This problem is corrected
by adding to the calculated time series 0 with the formula [0; diff(x1)]:
• cumsum (integration); cumulative sum of values
• smooth filtering in the desired window.

Collection of Formulas dialog

The Collection of Formulas dialog is a page opened from the Tools menu. This
page contains the most commonly used calculation formulas and general
instructions to create and edit calculated measurements. The Collection of
Formulas dialog includes examples of filters and formulas used to handle process
disturbances.

50
 Figure 9.1. Collection of Formulas dialog

In the Collection of Formulas dialog, you can click links to add measurements.

Adding calculated measurements from the Collection of Formulas

dialog to a diagram
To add a new calculated measurement for one or several measurements:

1. Select the necessary measurement(s).

2. Choose Collection of Formulas... in the Tools menu. The Collection of

Formulas dialog will appear.

3. Select a formula from the collection and click the Add calculated measurement
link. The Parameters dialog will appear. The text boxes show the name(s) of the
selected measurement(s) and default values set by the program.

4. Change the default values in the Parameters dialog as necessary.

5. Click OK.

51
 Figure 9.2. Parameters dialog

Adding calculated components

When a diagram is in the edit mode, you can add components to the diagram
canvas via the right-click menu.

Figure 9.3. Add new component menu

The New calculated measurement option adds a new calculated measurement.

By clicking this option, a component is added directly to the process diagram. To
edit the component, you must reselect it.

To add a calculated component to a diagram or subprocess:

1. Select the measurements that will be used in the calculation.

2. Choose a suitable location in the diagram for the new component.

3. Right-click the mouse and select New calculated measurement from the menu.

4. Right-click the component. From the menu that opens up, select the
Component properties option. The Measurement properties dialog is displayed

52
Figure 9.4. Calculated measurement, General tab

5. Choose the General tab. Enter:

- The component name in the Name: field

- The measurement unit in the Unit: field.

- Choose whether to display the component image and choose an icon for the
component.

- Select the position of the label.

- Choose whether to display the name of the measurement.

- Choose whether to display the last value of the measurement and the unit in the
diagram.

- If necessary, choose the time interval for the calculation of the value shown in the
diagram. This is useful if the calculation is, for example, the cumulative production
amount per day and if the calculation of the result requires the data of the whole
day.
- Use as state measurement - selection imports the the measurement to be used in
the Grade selection menu.

53
Figure 9.5. Calculated measurement, Formula tab

6. Choose the Formula tab. If you have selected two or more measurements, the
default value of the diagram is the sum of the variables. Edit and enter the new
calculated formula in the Formula field. The formula variables are expressed by
the names x1, x2, x3, etc.

Measurements can be added and removed from a calculation by pressing the Add
variables and Delete variables buttons.

You can change the order of the x1, x2 and x3 measurements with the Up and
Down buttons.

To select different measurements, click the name of the measurement to open a

drop-down menu.

7. Choose the calculation location: local/server.

8. Click OK. The Calculated measurement dialog will close.

54
10. Reference time
Reference time means that you can set a time interval from the history for use as a
reference for the present situation. This function is particularly useful when the
process has been changed and you want to know the changes in the measurement
values between before and after situations.

Figure 10.1. Set Time Intervals list box

Choosing reference times

To define a custom time interval for comparison:
1. Click the Set Time Intervals list box or the drop-down arrow on the right-hand
edge. A drop-down list will appear.

2. Click the Open Reference Time dialog… command. The Use Reference Time
dialog will appear.

Figure 10.2. Use Reference Time window

3. Select a reference time using either of the following two alternatives:

• - Use the current time interval as the reference time.
• - Use a favorite time interval as the reference time: select a favorite time interval
from the dialog.
4. Click OK to close the dialog. The reference time interval is activated in the trend

55
 Figure 10.3. Set Time Intervals list box when using a reference time

plots.

Use of the reference time is indicated on the Set Time Intervals shortcut bar, with
the start and end times of the trend plot time interval separated by three dots.

Figure 6.10 contains a trend plot whose left-hand part shows a Reference Time
Interval depicted with a green line. The right-hand part of the trend consists of new
data and is depicted with a black line.

Figure 10.4. Trend plot when using a reference time interval

Removing reference times

To deactivate a reference time from a trend plot:
1. Click the Set Time Intervals list box or the drop-down arrow on the right-hand
edge. A drop-down list will appear.

2. Click the Open Reference Time dialog… command. The Use Reference Time
dialog will appear.
3. Click Do not show any reference time.

4. Click OK; the dialog will close and the reference time will be deactivated.

56
11. Compensation of delays
Delays in the process often complicate the analysis of the process. In Wedge,
process delays can be compensated before analysis. Wedge's delay components
are implemented utilising the versatile calculation features of the software. These
features can be used to delay variables measured in different parts of the process
to a single point called the zero point. Chronological delaying of measurements
facilitates analysis of the cause-effect relations of the process.
In this chapter, you will learn the following:
• Creating delay models
• Defining delay component configuration parameters
• Using delay components in process analysis.

Delay model
A delay model is created by placing delay group components for all process
phases in a diagram. Place a delay component between the delay group
components and configure the type of delay that the delay component causes in
the process. Combine the components with a line. Line drawing direction and order
of connection of the delay components must match the actual flow direction of the
process, i.e., the direction of the delay.
Connect measurements belonging to each process phase to the delay group in
question. The background colour of the delay group's name will be copied to the
measurement name to help the user observe the delay model.

Figure 11.1. Delay model structure

The simplest method is to delay the measurement by a constant delay. This

works if the intermediate storage levels and flow rates of the process remain
constant.

If the flow rate is not constant, but varies, for example, in line with the production
speed, a pipeline delay should be applied. In this component, volume remains
constant but the flow rate varies.

If the volume of a given process part, such as a storage tower, does not remain

57
 constant, a tower delay should be used. In this case, the effect of changes in the
levels and flow rates of the intermediate storage towers on the delay time is taken
into account.

When the delay function is utilised, the measurements connected to the delay
group are delayed in accordance with the configuration of the delay components.

Building delay models

We will next describe how to build a delay model. Wedge contains the following
delay components:
• Delayed measurements
• Constant delay
• Pipe delay
• Tank delay
• External delay
• Join delay
• Split delay

The delay components are illustrated in the following figure:

Figure 11.2. Delay components

To add a delay component to a diagram:

1. From the Edit menu, select the Add components… command and the Delay
components tab.
2. Double-click the desired component. The program adds the component to the
diagram.

3. Correct the order of components in accordance with the process flow.

4. Connect the delay components to each other with a line in parallel with the
process flow.

5. To edit the delay component, right-click the component and select Component
properties from the menu.

Delay component properties

We will next describe the properties of each delay component and the parameters
specified for them.

58
Delayed measurements
The Delayed measurements component is used to give a name to a process
phase and to add to it the measurements belonging to the process phase in
question.
Use the component settings to specify the following:
• Component name in the Name: field
• Colour for the delayed measurements from the drop-down menu (12 colors to
choose from)
• Icon for the component
• Label position.

Figure 11.3. Delayed measurements, Configuration dialog

To add measurements to a delay group:

1. From the diagram measurement group, select the measurements that are
measured from the process phase in question.
2. Move the mouse pointer over the added Delayed measurements component.
Right-click the mouse and select the Add selected measurements to delay
group command.
3. The name fields of the measurement components selected become framed. The
frame colour is the same as the colour you selected for the Delayed
measurements component.
Whenever you move the mouse onto the Delayed measurements component, the
names of the measurements belonging to the delay component change to light
blue. This helps to identify the measurements that belong to the component in the
diagram. The next figure shows an example of a Delayed measurements
component and the measurements belonging to it.

59
 Figure 11.4. Delayed measurements component and measurements belonging to the
component

Constant delay
A constant delay component delays measurements by a standard time period.
This can be used if the intermediate storage levels and flow rates of the process
remain constant.

Use the component settings to specify the following:

• Delay: length of the constant delay
o Full days DD, maximum value 99 days
o Hours, minutes and seconds hh:mm:ss
• In the Icon and label tab, you can specify the following:
o Icon for the component
o Name
o Label position.

Figure 11.5. Constant delay component, Configuration dialog

Pipe delay
If the delay is not constant as a result of a non-constant production rate, you can
delay measurements using the Pipe delay component based on the constant
volume and the dynamically changing flow data.

Use the component settings to specify the following:

• Total pipe volume [m3]: enter the size of the process part in cubic metres.
• Pipe flow measurement: choose the name of the measurement from the drop-
down menu and the correct unit from the drop-down menu on the right-hand
edge.

60
 • In the Icon and label tab, you can specify the following:
o Icon for the component
o Name
o Label position.

Figure 11.6. Pipe delay component, Configuration dialog

Tank delay
If the surface level of process tanks varies, use the Tank delay component.

Use the component settings to specify the following:

• Total volume of the tank [m3]: enter the tank size in cubic metres.
• The part of the tank that is mixed [%]: estimate of the proportion of the tank
volume that mixes ideally.
• Tank level measurement [%]: select the measurement name from the drop-
down menu.
• Tank flow measurement: choose the name of the measurement from the drop-
down menu and the correct unit from the drop-down menu on the right-hand
edge.
• Inflow or outflow: click the check box according to whether the selected chest
flow is inwards or outwards.
• In the Icon and label tab, you can specify the following:
o Icon for the component
o Name
o Label position.

61
 Figure 11.7. Tank delay, parameter entry

The tank delay calculation is based on the total volume and surface level data of
the tank. The calculation gives the amount of material in the tank. This amount is
then divided by the flow data. Stock is assumed to progress as plug flow.

In practice, the tank delay is a combination of plug and mixed flow delays. Thus,
tower delay allows you to set the following:
• The proportion of mixed flow volume in the tower, the amount exceeding which
being plug flow, which directly affects the amount of delay
• The remaining portion in the tower, which is an ideal mix. If the surface level of
the tower is lower than the value defined for the mixed flow volume, then the
delay caused by this mixed volume is shortened.

External delay
The External delay component uses delay data obtained with calculated
measurements in Wedge or delay data calculated in another system.

Use the component settings to specify the following:

• Delay from measurement: select the measurement name from the drop-down
menu.
• Delay unit: select a unit from the drop-down menu.
• In the Icon and label tab, you can specify the following:
o Icon for the component
o Name
o Label position.

62
 Figure 11.8. External delay component, Configuration dialog

Join delay component

If the process runs parallel lines, the join delay and split delay components are
used. Information on when the stock is fed to each line is needed. This information
can be obtained, for example, from flow valve positions.

You can combine flow measurements in the diagram using the Join delay
component. Place the Join delay component to a suitable position on the diagram.
Connect two or more delay components with a line to the Join delay component.
Use the component settings to specify the following:
• Inwards flows combined with this component
• Measurement units of flow measurements from the drop-down menu.
• In the Icon and label tab, you can specify the following:
o Icon for the component
o Name
o Label position.

Figure 11.9. Join delay component

63
Split delay component
You can separate flow measurements in the diagram using the Divided flow delay
component. Place the Split delay component to a suitable position on the diagram.
Draw two or more lines from the Split delay component to the following delay
components.
Use the component settings to specify the following:
o Icon for the component
o Name
o Label position.

Figure 11.10. Divided flow delay component

Delaying Trend plots

Each Delayed measurements component that is added to a diagram also acts at
the same time as a zero point with which measurements can be delayed.

To delay measurements previously located in the process with a specific

measurement, right-click the mouse on the measurement in the Trend window and
select the Delay measurements command.

Once this is done, you can continue using the program as normal. Wedge
manages the temporal alignment of measurements with the zero point in question.
In doing so, the program takes into account all defined calculation points between
the actual measurement location and the analysed zero point. The delay
calculation is performed by consecutively linking each individual delay.

Once Wedge has automatically aligned the measurements with the selected zero
point, all basic Wedge functions, such as multivariate analysis (Principal
Component Analysis, PCA), can be utilised.

64
Adding and removing delayed user annotations
Once the calculation point (towers, washer) data has been configured with delay
components, the same information can also be utilised to monitor process
changes. To do this, select a time interval of interest from a given variable. The
progress and mixing of the stock during the time interval in question are expressed
in the trend plots of the measurements later in the process by means of
background coloring. The darker the background colour in the trend, the higher the
amount of selected stock flowing in the process during the selected time period.
If you select the Add delayed user annotation command, the software can
calculate the progress of the process status during the selected time interval. This
requires that the analysed measurements have been connected to a delay model.
The Delay measurements and Add delayed user annotation commands are
mutually exclusive. You can only use one property at a time.

To enable delayed background coloring in a trend:

1. Select the desired time interval from the trend.

2. Right-click the trend plot. Select Add delayed user annotation from the menu
that opens up.

Right-click the trend plot to deactivate the delayed background coloring. From the
menu that opens up, select Delete delayed user annotation. All areas marked
with delayed background coloring are removed.

Figure 11.11. Delayed user annotation

65
In the above figure, a Delayed user annotation has been marked in the
measurement "Pine chips to digester". As the stock flow progresses in the process,
the Delayed user annotation is visible in the measurement "Blow off kappa"
approximately 4 hours later.

66
12. Data analysis
This chapter presents the data analysis stages and the analysis tools:
• Data selection and processing
• Statistics
• XY plot
• Histogram
• Correlation matrix
• Cross-correlation
• Spectrum
• Periodic fluctuations
• Searching for fluctuation source (MAR)
• Example of MAR Navigator results
• Waveform matching.
Principal component analysis (PCA) is presented in full in a separate chapter.

Uses of analysis tools

Analysis tools can help find the source of a disturbance faster than traditional trend
tools. Analysis tools also offer information that trends do not show.

Various phenomena have their own specific analysis methods:

• Sudden changes are analysed using waveform matching.
• Constant changes are analysed by looking for the fluctuation source using MAR
analysis.
Wedge's analysis tools can be accessed through the Analyses menu or by clicking
the More functions button in the open Trend window. Analyses are based on
measurement data. They are session-specific, that is, they are not permanently
saved. Analyses are not usually performed on all measurements in the diagram at
once; rather, the user selects a certain number of measurements for which a
suitable analysis is then performed. The number of measurements that can be
selected depends on the analysis method used.

The analysis results provided by Wedge are calculated for the measurement data
selected by the user. Wedge's analysis methods are advanced and highly
automated. However, the measurement data used for analysis must be carefully
selected to safeguard the reliability and usefulness of the results obtained. In
addition, the selected measurement data often requires preprocessing. The user
must, for example, decide whether or not data recorded during a production break
or shutdown should be included in the analysis.

Pre-analysis procedures include the following:

• Setting analysis time intervals
• Setting time levels

67
 • Examining data as trends
• Preprocessing data.

The analysis can be focused on a specific process state by using state

dependency.

Selecting analysis time intervals

The choice of time interval for the analysis and, consequently, the analysis method
used depends on what you wish to examine. Instructions for selecting suitable time
intervals are provided in connection with each analysis method presented in this
User Manual.

As a general rule, the selected time interval must be long enough, and the
phenomenon under analysis should occur as strongly as possible during it. When
determining the cause of a periodic fluctuation, the examined fluctuation must
occur during the selected time interval as strongly as possible while, on the other
hand, the selected time interval itself must also contain at least several cycles of
variation. When identifying the cause of individual disturbances, process delays
must also be taken into account. The cause of a disturbance can often be
temporally far removed due to lengthy delays. On the other hand, too long a time
interval can also be detrimental. In such a case, a phenomenon may be identified
that closely resembles the analysed phenomenon, but is so far off timewise that it
could never have brought about the analysed phenomenon. In turn, when selecting
the reference period for statistical monitoring of the process, one must remember
that the process must be in the normal state during the reference period.

Setting time levels

The time level (sampling interval) selection for the analysis depends on the length
of the history used, the speed of the analysed phenomenon and process delays.
The selected time level should be fast enough to enable the analysis method to
perceive the delay between the cause and the effect. In addition, at the selected
time level, the analysed phenomenon must be the most clearly apparent and all
other phenomena the least apparent. As a general rule, the time level must be at
least twice as fast as the fluctuation examined. It should also be noted that the
values within the sampling interval are treated as simultaneous.

An example of the effect of the time level on the phenomena observed from the
data is shown in the figures below. The trend plots present the same data, but in
the latter trend plot the time level is 30 times slower. Furthermore, the spectrum
diagram calculated from the latter measurement data provides faulty data
concerning the periodic nature of the measurement.

68
 Figure 12.1. Fast time level in trend plot

Figure 12.2. Slow time level in trend plot

The selected time level should be fast enough to enable observation of the
examined phenomenon. On the other hand, too fast a time level is detrimental
because the amount of data increases to undesirable levels. If, for example, you
are analysing pH variation, a 10-second time level will be far too short. If you are
unsure of the suitable time level, too fast a level is, however, always preferable to
too slow.

Preprocessing data
Measurement data can contain data that is faulty or otherwise unsuitable for
analysis. To ensure reliable analysis results, such unsuitable data must be either
deleted or repaired before commencing the analysis.

69
 The definition of data suitable for analysis depends on the analysis method used.
For example, when clarifying the cause of a non-recurring change using the
Waveform matching method, tuning the process measurement devices is not
essential. If, on the other hand, a balance calculation is carried out using
Calculated measurement, the results obtained will be incorrect if the tuning is not
right. There are no set, universally applicable instructions for selecting data and
carrying out the required data preprocessing. This User Manual does, however,
provide instructions for each of the presented methods regarding key factors to be
taken into account when applying the different analysis methods.

Stationarity of measurement data

Wedge uses statistical and time series analysis methods to identify the cause-
effect relations of process disturbances and to locate the causes of process
disturbances. When statistical methods are applied, obtaining reliable results often
requires measurement data stationarity. Data is stationary if its statistical
properties, such as average and deviation, are not time-dependent.

The most common non-stationarities of measurement data are:

• Trend, i.e., slow variation of the average. Trends can be identified by visual
examination of the measurement data. A trend can be brought about, for
example, by low frequency drifting in the process or measurement sensor
creep. To delete a trend, the Trend removal calculated measurement in
Wedge's collection of Matlab formulas can be used.
• A Step change can be created, for example, when the process running
conditions are modified by changing a set point. Such a change can also be
caused by a process event, such as a device error. A step change can usually
also be observed by visual examination of the measurement data.
• Peaks are single measurement values that statistically deviate highly
significantly from other measurement data. They can be identified by visual
examination of the measurement data. Peaks can indicate a sensor fault, or
they can be actual values, in which case they indicate a process disturbance.
Interpretation of the causes of peaks requires knowledge of the process.
Statistical and time series analyses of measurement data containing peaks can
easily produce faulty conclusions. More reliable analysis results are obtained by
ruling out the most inconsistent values, in other words, by using data screening.
• Temporal variations in fluctuation strength. Sometimes the strength of
fluctuation of a process variable can change fundamentally, and yet the average
fluctuation remains unchanged. Such an event indicates a process state
transition and, as such, the data is not stationary.

It is often difficult to find within measurement data a long enough time interval that
is free of the above non-stationarities. In such a case, the measurement data can
be preprocessed to make it better suited for analysis.

Missing and faulty data points

Disturbances sometimes occur during measurement data collection, whereupon
the data points either go missing or their values are clearly faulty (so-called
outliers). Missing data points can be caused either by a single random data
collection error, or they can occur for a longer time interval as the result, for

70
 example, of a sensor fault. Wedge's analysis methods take into account missing
data points, and they are calculated for in the most suitable way for each method.
Faulty data values can be caused, for example, by measurement errors or
abnormal conditions. Faulty values can also occur over long time intervals, for
example, for the duration of a production stoppage. If faulty values are not
separately taken into account in the calculation, they will distort the analysis results
calculated from the measurement data.

Data preprocessing methods in Wedge

The purpose of data preprocessing is to convert the measurement data into a form
that can be used to obtain reliable analysis results. In Wedge, faulty measurement
data can be handled either by cutting the data out or by defining the data as
missing.

Data cutoff
The easiest and most common method of data preprocessing is to cut the intervals
containing faulty measurement data or no valuable measurement information from
the measurement data. Measurement data can be cut (data cutoff) by either
removing the time interval in question or cutting off all intervals during which the
selected measurement has exceeded the set upper or lower limits. The data cutoff
is applied to all diagram measurements. Detailed instructions for data cutting are
presented in the chapter Viewing trends, Removing a time interval that exceeds
limits and Removing all time intervals that exceed limits.

Data cutoff causes discontinuity of the time axis. However, Wedge's calculations
are carried out so that the cut data does not distort the obtained analysis results.
Nevertheless, it must be noted that important information can be lost when cutting
data from other measurements.

Saving cut time intervals as a Favorite time interval

If you want to save an examined time interval for later viewing, you can save it as a
Favorite time interval. For more instructions, see the chapter User-specific
settings. The cut intervals are saved as favorite time intervals – you can reuse the
exact same cut time interval and measurement data in later analyses.

The saved Favorite time interval can be reaccessed via the Calendar view time
interval shortcut of the Trend window.

Automatic data cutoff

In Data screening, any clearly faulty data points are automatically removed before
performing the analysis. Data screening is applied to all diagram measurements. It
is based on a certain measurement called mask signal, which indicates the
process state. A mask signal can consist of machine running data or a production
stoppage signal. To perform data screening, select Measurement data screening
in the Options menu. Select Data screening in the dialog that opens up on the
display to set the constraints for data screening.

71
 Figure 12.3. Automatic data screening

Set the constraints by entering the name of the measurement on which data
screening is to be performed in the Measurement field. Next, select the
comparison operator from the drop-down menu. Finally, in the right-hand field,
enter the limit value with which the selected measurement is to be compared. Click
OK. The data you have classified as unreliable will now be screened from all
analyses. It is important to ensure that the data defined in the Data screening
dialog consists of the values to be accepted for analysis. Therefore, the values
which do not fulfil the constraint are screened out. In this example, the time
intervals during which the MASK measurement value is higher than 0.5 are
accepted for analysis.

Once Data screening is in use, the selection will be observed in all analyses and
the application will screen all data classified as unreliable by the user. Data
screening is a session-specific function, which means that the selection you have
made expires when you end the current session.

Defining data as missing data points

Measurement-specific data screening is applied when a measurement is
considered to contain faulty data but the process has otherwise maintained normal
state. By using measurement-specific data screening, no useful data is screened
from any other measurements.

Measurement-specific screening is carried out using a calculated measurement.

Measurement-specific screening can be based, for example, on the crossing of
different rationality limits (limit value or change speed) or too many consecutive
occurrences of a same given value.

To perform measurement-specific data screening, first select a measurement to

which screening is to be applied. For the screening, use Wedge's wscreen
function and enter the measurement for which data screening is to be performed
and the constraint(s) by which the values acceptable for analysis are to be
determined. If there are several constraints, use the operators & (AND operator) or
| (OR operator) to link them. The following figures shows an example of the use of
the wscreen function.

72
 Figure 12.4. Measurement-specific data screening using calculated measurements

In the example, data screening is applied to the measurement Hardwood HD cons.

In order for the data point of this measurement to be accepted for analysis, the
following constraints must be met:
• Hardwood HD cons > 4.5 and
• Hardwood mixing flow > 50.

If a data point meets the constraints, it will be kept. If either or neither of the
constraints is met, NaN (Not a Number) will be marked as the data point value.
Thus, the data points are not "cut out" as this would cause discontinuity within the
time axis of the data analysed. In this way, the data points considered unreliable
are screened from the measurement and are accounted for in the most suitable
way for each analysis method without lessening the reliability of the analysis
results.

Selection of state dependency

The analysis can also be applied to a specific process state, such as a produced
product. To be able to use state dependency-based functions, the diagram must
include a state-indicating measurement that represents the product currently
produced.
The process state can also be determined in ways other than on the basis of the
grade produced. This can, for example, be based on different performance
indexes, in which case, for example, the process states can be defined as good

73
 and poor production runs.

Analysis tools
Analysis methods are used to effectively identify the cause of process disturbances
and to obtain process information that would otherwise be difficult to obtain using
standard trend tools.
The various applications of Wedge's analysis methods are presented below in
brief. The aim is to help the user to select a suitable method for the analysis and
identification of the cause of different process problems.
• Statistical Process Control (SPC) is used to monitor and identify causes of
changes in process operations.
• Examining the Statistics gives a better picture of the measurement data and
the different process situations. The statistics function can be used, for
example, when studying the measurement data deviation or fluctuation range.
• The XY plot function can be used to examine dependences between two
measurements both graphically and between measurements using a derived fit.
• The Histogram function facilitates analysis of the measurement data
distribution.
• The correlation coefficient describes the linear relationship, or similarity of
form, between different measurement data. It does not take into account delays
between measurements.
• The Best correlations analysis is used to search the measurements that best
correlate with the target measurement.
• Cross-correlation is useful for determining the delay correlations and auto-
correlation of measurements.
• The spectra are used to interpret measurement signal cycles and to display the
division of measurement signal effect on different frequencies.
• Periodic fluctuations analysis is used to identify fluctuations of different
measurements that occur at the same cycle length.
• The MAR Navigator is used to find the origin of periodic process fluctuations.
• The waveform matching is used to determine the cause or consequences of
non-recurring phenomena.
After an analysis has been made, use the handy More functions menu to continue
with the process. Use the More functions menu of the open statistics dialog, for
example, to select a correlation matrix to view the correlation coefficients of the
measurements in question.

Studying groups of measurements in the trend window

It is recommended that you study the data before conducting a data analysis to
discover and remove faulty values. For quick study of trends, use the Compare
trends with this measurement function. This command means that two trend
plots are shown at the same time in the opened Trend window. The topmost trend
plot is the selected measurement with which other trend plots are compared. Use
the command as follows:

74
 1. Choose one or more measurements from the diagram.

2. Hold down the right mouse button over the measurement to be set as the
reference. Select the Compare trends with this measurement command from
the menu that opens up.
• If you have selected one measurement only, all diagram or measurement group
measurements will be compared with this measurement.
• If you have selected more than one measurement, the other selected
measurements are the reference.
3. To browse the trends, use the scroll bar that appears on the right-hand edge of
the lower trend plot.

In the next figure, the measurement Hardwood mixing flow, with which all other
measurements of the diagram are compared, has been selected. There is a scroll
bar on the right-hand edge. At the bottom of the figure, you can see the More
functions button. Use the button to select analysis functions to be applied to the
data.

Figure 12.5. Compare trends with selected measurement

While browsing the trends, you can easily screen any unwanted time intervals from
the analysis to improve the reliability of analyses.

Statistical process control (SPC)

The statistical analysis monitors whether the signal values statistically differ from
the reference period. The reference period is a time interval selected from the
history with which you can compare the process state at a later time. On the basis

75
of the reference period, the alarm limits are defined within which the measurement
values must remain in order that the process can be considered to be in the same
state as it was during the reference period, that is, the measurement can be
considered to be statistically under control.

Correct selection of the reference period is vital to the success of the statistical
process analysis and requires good knowledge of the process. During the chosen
period the process must be in the normal state. There must be no exceptional
fluctuations and the quality must be optimal. In order to ensure that the reference
period is truly representative, the selected interval must be of sufficient length.

SPC works on the basis of set alarm limits that are defined on the basis of the
chosen reference period. The reference period is selected on the basis of process
knowledge. All time period selection properties of Wedge can be applied when
selecting the reference period. Data cutoff can be also be used. State dependency
should also be observed when using statistical process control. If the time interval
contains different states (such as the end product grade), the statistical reliability of
the limits calculated on the basis of the interval cannot be guaranteed. In that case,
the limits must be defined state-specifically. If state dependency is used for
defining limits, it is recommended that several periods during which the process
has been in the same mode are included in the reference period. For more detailed
information on setting limits, see the chapter Special trend functions.

You can also set alarm limits based on your own knowledge of the process. The
limits can also be set asymmetrically.

Figure 12.6. Trend plot showing alarms, target value and upper and lower limits

Once the limits have been defined and Alarms has been selected in the
Graduations dialog, all intervals during which the measurement has fallen outside
the set limits are displayed with a yellow background in the measurement trend
plots. If Target value is selected, it is shown in the trend plot in green and if Upper
limit and Lower limit are selected, they are shown in red. When Calculate from
current data – average is selected in the Graduations dialog, the average is
shown in the trend plot in blue.

76
Statistics
The Statistics command displays the statistical indicators of the measurement
data as a table. By pressing the Columns button, you can select which of the
following pieces of statistical information will be displayed as columns:
• Unit
• Tag Average
• Deviation, indicates how much the observed values on average deviate from
the observed average
• Relative deviation, gives the significance of the deviation. Deviation alone
describes the absolute fluctuation. Relative deviation can, for example, be used
to compare measurements from different measurement ranges.
• Maximum, the highest value of the each measurement in the examined time
interval
• Minimum, the lowest value of the each measurement in the examined time
interval
• Fluctuation range (maximum–minimum)
• Last value, the last measurement values in the examined interval
• Number, the number of data points used in the statistical calculation.
The number of measurements that can be selected for statistical analysis is not
limited.

Statistics can be searched grade-specifically. To view the statistics for an individual

grade, select the name of the grade from the Select Data list box. The values of
the Statistics dialog are calculated only from the data points belonging to the
grades in question. You can also select several different grades for simultaneous
viewing.

Figure 12.7. Statistics dialog

Click the name of the measurement to open the trend plot of the measurement.
Move the mouse pointer over the name of the measurement to show a red check

77
 mark that allows you to remove the measurement from the table. You can drag
new measurements from a diagram to the statistics dialog.

Click the column name to arrange the statistics table by the data in the column in
question (ascending or descending).

XY plot
The XY plot can be used to study the dependency between the simultaneous
values of two measurement signals. This kind of analysis can be performed, for
example, when estimating the reliability of the calculated correlation coefficient
between two measurements (see the chapter Correlation coefficient). Two
measurements must be selected for the XY plot.

Figure 12.8. XY plot dialog

A 1st-, 2nd- or 3rd-degree plot can be applied to an XY plot. The equation of the
fitted plot and the r2 value are shown at the bottom of the XY plot dialog. The
degree of the plot to be fitted can be changed via the drop-down menu on the
bottom right-hand edge of the canvas. The axes of the measurements can be
changed via the button next to the drop-down menu, upon which the fitted equation

78
 is recalculated.

The time axis of the measurement points is presented in colour. Sampling points
drawn in blue represent data points that date further back in the history, while red
data points are more recent.

XY plots can also be drawn grade-specifically. The details of data drawn on a

diagram can be examined using the zooming function. To zoom in, first left-click on
the canvas. Then, holding the left mouse button down, drag a box around the
section of the plot you want to zoom. Next, right-click on the plot. From the menu
that opens up, select Zoom in.

To return to the original plot view, right-click on the plot. From the menu that opens
up, select Zoom out.

The XY plot can be copied and, for example, pasted into a Microsoft® Word
document. To copy the plot, right-click on it. From the menu that opens up, select
Copy graphics. You can then paste the copied plot into a Word document via the
clipboard.

Histogram
The Histogram function shows the number of measurement points in each value
range of the measurement. It can be used to show the measurement deviation
(frequency function) as a bar chart and the cumulation (cumulative function) as a
curve. A histogram is a standard tool used to present information contained in the
large amounts of data. Its X-axis is the value range of the measurement result and
the Y-axis is the rate of occurrence of the measurement result.
The histogram shows the minimum and maximum values and fluctuation range of
the measurement data and can be used to estimate average and deviation values.
The histogram shape can also be utilised when analysing the measured data. The
shape of the histogram is used when estimating the proximity of the measurement
data distribution to the normal distribution. The shape can also help to determine
whether the measurement results are representing the same phenomenon. For
example, a grade change can lead to the occurrence of a so-called isolated peak.
Analysis of measurement data of this type on its own can give misleading results.

The cumulation plot of the histogram reveals the cumulation of measurement data,
in other words, it describes how the observations accumulate as the value of the
variable increases. The cumulation clearly shows what proportion of observations
lie above or below a given limit.

79
Figure 12.9. Histogram dialog

Figure 12.10. Histogram's Cumulation dialog

To open a histogram in the cumulation view, select Cumulation in the Histogram

dialog. To return to the normal view, deselect the Cumulation option. You can
select one to ten measurements at a time for a histogram.

The histogram window can be copied and, for example, pasted into a Microsoft®
Word document. To copy the plot, right-click on it. From the menu that opens up,
select Copy graphics. You can then paste the copied image into a Word
document.

80
Correlation matrix
The correlation coefficient describes the linear relationship, or similarity of form,
between different measurement data. The coefficient ranges between -1 and 1.
The higher the absolute value of the coefficient, the closer the linear relationship
between the analysed measurement data. When the linear relationship is perfect,
the correlation coefficient is either -1 or 1. When the correlation coefficient is
positive, the linear relationship is the same, in other words, as the values of one
variable increase, the values of the other variable also increase. When the
correlation coefficient is negative, the relationship is the inverse, in other words, as
the values of one variable increase, the values of the other decrease. The
correlation coefficient takes no account of delays in the system; rather, the
calculation is based on simultaneous values of different measurements. Delays can
be addressed by means of Cross-correlation. Process delay refers to a time
delay with which the correlation coefficient between measurements achieves its
absolute maximum value.

It must be noted that the correlation coefficient indicates a linear relationship. The
correlation coefficient can be zero, even if the variables have a statistical
relationship. In such a case the relationship is non-linear. It is also important to
note that even a single deviating observed value strongly affects the correlation
coefficient. The correlation coefficient can be close to zero even if most
measurement data seems to correlate strongly. For this reason, it can be useful to
first draw, for example, an XY plot. The plot will show whether the deviating values
affect the correlation coefficient. Here, again, it is important to remember that the
correlation coefficients reveal nothing about the cause-effect relation, the
estimation of which is based on knowledge of the process. It should also be noted
that process control feedbacks can mean that the measurement derived from the
earlier process stage is not necessarily the cause of the phenomenon in question.
The correlation matrix shows the correlations between all selected signals. All
values of the correlation matrix diagonal are 1, because the signal's correlation with
itself is always perfect.

The number of measurements that can be selected for a correlation matrix is

unlimited. However, if many measurements are selected for analysis, the
performance time can be lengthy. In addition, the outcome dialog can become so
large that it no longer fits fully on the display. In the correlation matrix window,
correlations are displayed both numerically and in colour. A blue colour indicates a
positive correlation and red a negative correlation. The lightness or darkness of the
colour indicates the weakness or strength of the correlation, respectively.

81
 Figure 12.11. Correlation matrix dialog

Click the name of the measurement in the first column to open the trend plot of the
measurement. Move the mouse pointer over the name of the measurement to
show a red check mark that allows you to remove the measurement from the table.
You can drag new measurements from a diagram to the correlation matrix dialog.

Click the name of the measurement in the column title to sort the correlation
coefficients in the column by their absolute values. In the window, you can also see
the trends for the selected measurement and the trends for the measurement that
correlates best with the selected measurement. To change this lower trend, click
the name of any other measurement in the list on the right-hand edge of the dialog.

Figure 12.12. Correlation order dialog

Best correlations
In the previous chapter, we explained that you can sort the measurements of a
correlation matrix by the absolute value of correlation coefficients by clicking a
correlation matrix column name. You can also do this by opening a trend of any
measurement and selecting the Best correlations tool from the More functions
menu. In this case, correlation coefficients are calculated for all the measurements
of the measurement group in use or for the selected measurements.

82
Cross-correlation
Cross-correlation can be used to study the correlation between two measurements
with various delays. Cross-correlation calculates all possible delays up to a given
limit, so the user does not need to know all delays between the signals in advance.
Auto-correlation, in turn, refers to correlation within measurements. Auto-
correlation is a time separation function that indicates the degree of correlation
between signal data points that are separated by a given time interval.
The correlation coefficient between signals is shown at the 0 point (no delay) on
the time axis of the cross-correlation plot. If the correlation function shows a single
clear minimum or maximum, the extreme value point indicates the delay between
the signals, and the correlation function value at this point indicates the value of the
delay-compensated correlation coefficient. When calculating the cross-correlation,
it should be noted that process control feedbacks can mean that the measurement
derived from the earlier process stage is not necessarily the cause of the
phenomenon in question. If the correlation function has one minimum and one
maximum value, one being at a negative point of time and the other at a positive
time, these usually indicate an incoming connection via the process and an
incoming connection via the control. The process control delays can be identified
from these extreme value points.
Two measurements must be selected for cross-correlation. A cross-correlation plot
for the selected signals is shown, together with a table showing the contribution
direction, the strongest correlation and the delay with which the strongest
correlation is obtained. If no maximum or minimum is found, the table shows no
max/min values.

Figure 12.13. Cross-correlation dialog

If only one measurement is selected, an auto-correlation is obtained.

83
To zoom in on the Cross-correlation dialog, first left-click on the canvas. Then,
holding the left mouse button down, drag a box around the section of the plot you
want to zoom. Next, right-click on the plot. From the menu that opens up, select
Zoom in.

To return to the original plot view, right-click on the plot. From the menu that opens
up, select Zoom out.

The Cross-correlation dialog can be copied and, for example, pasted into a
Microsoft® Word document. To copy the plot, right-click on it. From the menu that
opens up, select Copy graphics. You can then paste the copied image into a
Word document.

84
Periodic fluctuations
A spectrum is a graphical process data display format that is used to interpret
measurement signal cycles and to display the division of measurement signal
fluctuations at different frequencies. It can be used to distinguish simultaneous
phenomena according to cycle length.
They are especially useful when the user notices in a trend plot that the
measurement data has periodic characteristics. In addition, weak periodic
fluctuations can be visualised using a spectrum.
Periodical characteristics are described in terms of interval time, frequency and
amplitude.
The interval time (T) refers to the length of a repeating cycle. Frequency (f) is the
inverse of the interval time. The frequency unit is hertz (Hz) and frequency is
calculated using the following formula:

f = 1/T
Amplitude refers to signal height as measured from the average to the highest
point.
Spectra are usually used to identify general trend forms as opposed to precise
numerical values. Spectra can give information on the interval time and strength
(effect) of periodical components. They also help to differentiate weaker periodical
fluctuations that are easily ignored when viewing a time sequence that contains
stronger fluctuations. Periodic fluctuations are shown in the spectrum as peaks or
waves.
A process can contain several simultaneous fluctuations, such as resonating
control circuits, regulating unit malfunctions and raw material variations. Examples
of the origins of periodic fluctuation:
• Fast: fluctuation caused by a rotating engine or a pump with a defective bearing
• Slow: vibrating control circuit
• Broad wave in a spectrum: raw material variation
The number of measurements selected for period fluctuations analysis is not
limited, but if more than ten measurements are selected, spectrum diagrams
cannot be used. The Periodic fluctuations command displays a spectrum
window. The area under the curve describes the fluctuation effect. Periodic
fluctuations are shown as peaks or waves. The table next to the spectrum diagram
shows the cycle length and the effect of the fluctuation on that period as a
percentage of total fluctuation.
The fluctuation source can be found using the MAR Navigator. The MAR
Navigator can be started directly via the table next to the spectrum diagram by
selecting the desired fluctuation period strength percentage with the mouse.
Alternatively, you can first examine the measurement spectrum diagram and then
manually select the desired fluctuation period.

85
 Figure 12.14. Periodic fluctuations dialog

Opening the Spectrum dialog

You can also open a separate Spectrum dialog. You can view and edit a spectrum
containing one or several measurements. The selected spectra are displayed
either separately or together in the Spectrum dialog.

To view the spectra for measurements:

1. Select the measurements from the diagram.

2. Right-click the diagram and select Separate Spectra or Combined Spectra.

Figure 12.15. Spectrum dialog

Spectrum dialog
The Spectrum dialog shows spectra in a spectrum plot. The boxes and buttons of
the Spectrum dialog function similarly to Trend window buttons, but the
Properties list box has the following three time axis formats:
• Frequencies, expressed numerically as 0-XXX

86
 • Seconds, expressed as 1e10-XXX
• Hours, expressed as HHHH:MM:SS – H:MM:SS.

Figure 12.16. Spectrum dialog's Properties menu

When interpreting a spectrum, it is often advisable to rule out the slowest

frequency, because if the slowest frequencies are displayed, the most informative
frequency peaks can easily be scaled out. If you want to keep the slower
frequencies, you can apply a logarithmic Y-axis by checking the Show logarithmic
values check box in the Trend settings dialog. This enables you to scale the
peaks of different heights of the different frequencies on the same display.

Searching for fluctuation source (MAR)

Periodic fluctuations in processes can result, for example, from poorly tuned
controls, malfunctioning valves, variations in raw material quality or daily or
seasonal variations. A process often contains several simultaneous disturbances. It
is common to controlled systems that a given fluctuation is often manifested at
several points throughout the process. Process fluctuations can, for example,
reduce productivity and lower the quality of the end product.
The MAR Navigator tool is used to determine the origin of process fluctuations. The
MAR Navigator is based on multivariate auto-regressive modelling. Automatic
processing of system feedbacks is the key attribute of the tool.

Theory of MAR modelling

MAR modelling is based on the use of measurement signals received from the
process to carry out model identification (so-called Black Box modelling). MAR
takes into account system feedbacks and the different routes of effect of process
disturbances. In MAR modelling, the measurement value x i is explained using
earlier measurement values from the same measurement and other
measurements.

87
 m
x t =  a(l)  xt – l + et 
l=1

The vector x[t] contains the values of the measurement signals at the time t.
According to the model, the xi (i=1, 2, …, n) value of a given signal at time t
depends on its past values and the past values of other signals up until the
moment t-M. The order (m) of the model indicates how many sampling interval
length time steps backwards in the process history have a current influence in the
model. n is the number of measurements included in the model. Matrices a(1),
a(2), …, a(m) are (n x n)-sized weighting matrices. The weighting factors are
identified by forming step responses between all variables using the least squares
method. The factors describe the interactions between the signals. They show the
effect of past values on the signal value at each delay 1, 2,…, m.

The vector's e[t] elements are referred to as noise sources. They describe the
share of the fluctuation that cannot be explained using the earlier values of the
measurements selected for the analysis. For example, variations in raw material
quality and flow fluctuations caused by malfunctioning valves cannot be explained
using earlier values, so these phenomena are shown in the noise sources e[t]. The
sources of disturbance located near the xi value of the corresponding
measurement that are causing process fluctuations are visible in all noise sources.

Using the mathematical form of the MAR model, the fluctuation of each
measurement can be presented as the sum of the effects of the noise sources.
Each noise term's share of the analysed signal is referred to as noise contribution
and is typically expressed as a percentage. These percentages indicate what
proportion of the fluctuation of the analysed measurement originates from the
sources of disturbance located near each of the measurements included in the
model. The highest percentages therefore show which measurements are closest
to the fluctuation sources, and thus where the search for their origin should be
focused.
The measurement data of the measurements selected for the MAR model must not
be too similar. If so, the modelling function is unable to distinguish between the
different possible progress routes of the process fluctuation and the model
identification will fail. The most common causes of failure are the following:
• The value of a given measurement changes only a few times or not at all.
• The fluctuation of two different measurements is otherwise highly similar, for
example, measurements of the same variable that are located close to each
other.
• The model includes both a calculated measurement and the measurements
used in its calculation; especially when the calculation formula is linear as, for
example, in the calculated measurement "3*x1" or "x1+x2".

Because the error is caused by the similarity of the fluctuation of the measurement
data selected for the modelling, the measurements to be included in the analysis
must be reselected, ensuring that the aforementioned similarities between
measurements do not occur.

88
 Processes often contain numerous simultaneous fluctuations. For example, poorly
tuned control loops often cause fluctuations at specific cycle lengths. MAR analysis
is, in fact, usually carried out at frequency level, so fluctuations occurring at
different frequencies can be distinguished from each other.

Although the MAR model is linear, the least squares fitting method it uses does
also go some way towards describing non-linear dependences. However, complex
non-linear dependences cannot be determined using MAR modelling.

For additional information on MAR modelling, refer to the following source:

Saarela, O., Multivariate Autoregressive Analysis in Locating the Origin of

Fluctuation in Continuous Industrial Processes, doctoral thesis, Tampere University
of Technology, 2002.

MAR Navigator
In Wedge's MAR Navigator, all of the complex MAR modelling calculations stages
are automated. The MAR Navigator results are presented as contribution
percentage values indicating the proportion of the overall target signal fluctuation
that is caused by each explanatory factor within the analysed time interval. The
percentage values obtained from the MAR Navigator are mainly intended to
provide a measure of the scale of significance of the different sources of
disturbance.

When using the MAR Navigator, the following points should be taken into account:
• Least squares fitting always provides some form of fit. Random deviations and
non-linearity can, however, cause fitting errors. The effect of random errors can
be reduced by using a large amount of data and the effect of non-linearity
reduced by using a fixed process state.
• MAR Navigator is designed to facilitate the analysis of dynamic process
fluctuations. The MAR analysis is unable to process static phenomena or
extraordinary transient phenomena.
• Successful selection and preprocessing of data is essential to achieving reliable
analysis results.
• The MAR Navigator automatically deducts the average from the time series and
removes the trend.
• As a Wedge user, you do not need to determine the model order as this
function is automated in the MAR Navigator.

It is important to note that a measurement does not necessarily exist for all
possible disturbance sources. For this reason, the resulting list of measurements
does not necessarily indicate the original cause of the disturbance; rather, where
the effect of the disturbances is most evident. In this way, the navigator locates the
process part where the original cause of the disturbance is most likely to be found.
Furthermore, the MAR Navigator can provide the variable's contribution percentage
for the target measurement of the analysis. This indicates whether the source of
disturbance is near the signal itself or whether the contributions are other sources
of disturbance not included in the analysis.
Processes often contain periodic fluctuations that continue only for a certain period
and may or may not be repeated later on. In such cases, only the time interval

89
 during which the fluctuation was observable should be included in the analysis.
The contribution percentages indicate the significance of the source of disturbance
within the data range selected for analysis. If a long range in which fluctuation
occurs only for a short period is analysed, the contribution percentages calculated
can easily be extremely small, making it difficult to determine the origin of the
fluctuation.

In MAR modelling, the fluctuation of signals is explained using the values of past
signals. The results obtained with the MAR Navigator are therefore unreliable if the
analysis contains many signals for which changes occur exactly simultaneously
(within a sampling interval). Simultaneous changes should therefore be examined
using correlation coefficients.

The MAR Navigator performs for the user an invisible reliability analysis, on the
basis of which each analysis variable is either rejected or accepted.

Using the MAR Navigator

In order to begin using the MAR Navigator, you must first select the measurement
for which you wish to determine the fluctuation origin. A Periodic fluctuations
analysis will be performed on the selected measurement. Then, in the table, click
the fluctuation strength percentage value for the desired cycle length.

The MAR Navigator dialog will appear. The time scale corresponding to the cycle
length is automatically shown in the Fluctuation time scale field. If you want to
restrict the analysis to a certain frequency, click Show spectrum to open the
spectrum diagram for the selected measurement. Select the desired area in the
spectrum using the X-limiter and then click Spectrum Area in the MAR Navigator
dialog.

Figure 12.17. Selecting a fluctuation frequency range in a spectrum

The time scale of the selected area is shown in the MAR Navigator dialog. The
time scale can also be entered directly in the MAR Navigator dialog's Fluctuation
time scale field.

90
Figure 12.18. MAR Navigator dialog

In the MAR Navigator dialog you can also choose whether you want to perform
the search for all process diagram measurements (Exhaustive search) or for the
selected measurement group only (Group-specific search…). If the exhaustive
search option is selected and the diagram contains a large number of
measurements, the analysis may take a long time to complete. If you want to
search for the cause of the disturbance in a specific group elsewhere in the
process, you should first create a measurement group. The group should contain
the target variable and the measurements for which you want to determine the
proportion of the target variable fluctuation. Once this is done, start the MAR
Navigator by selecting the created measurement group as the analysis target. This
speeds up the calculation compared to running an exhaustive search if the diagram
contains a large number of measurements.

The MAR Navigator displays the results of the analysis in the dialog. The result
includes the time scale of fluctuation and the analysed time interval as well as the
variables located in the vicinity of the identified origin of the fluctuation. Beside
each measurement is displayed the proportional contribution of the measurement
to the fluctuation as a percentage value. Clicking the Measurements button opens
a dialog showing which of the variables included in the analysis have been rejected
and on what basis, as well as the variables located in the vicinity of the fluctuation
origin. In addition, Wedge creates a temporary measurement group containing all
of the most significant fluctuation causing measurements. You can continue your
data analysis using the temporary measurement group. After the analysis has been
completed, Wedge also opens a Trend window which shows the target variable
and the trends of the variables identified as the most significant causes of its
fluctuation.

If the MAR Navigator finds the origin of the fluctuation to be a measurement that is
already known to you, repeat the analysis with the measurement in question
excluded. This can be done, for example, by removing any well-known sources of
disturbance from the selected measurement group. This will facilitate the
identification of new, unforeseen sources of disturbance.

91
 Figure 12.19. MAR Navigator Outcome dialog

The MAR navigator outcome dialog shows the following result: in this case, no
significant fluctuation source was found elsewhere. Either the disturbance
originates from the target measurement or no measurement is associated with the
fluctuation origin.

Example of MAR Navigator analysis in the food industry

The primary objective of the example process is the elimination of raw material
impurities. A key operational indicator of the process is process yield. Low yield
indicates that the raw material is not being utilised to optimal effect. On the other
hand, strict quality requirements regarding the impurity content of the end product
places an upper limit on yield.

Figure 12.20. Example food industry process

Yield fluctuations were a familiar problem at the plant. Yield had to be kept low in

92
order to ensure that the set limit on end product impurities was not exceeded. The
following figure shows a yield trend. Eliminating yield fluctuation would enable the
maximum yield level to be raised and would also enable more efficient utilisation of
valuable raw materials.

Figure 12.21. Yield Trend diagram for the example process

The fluctuation passed through several control circuits and return flows. All critical
process measurements were examined. The same fluctuation was found to be
present in the majority of measurements.

The yield spectrum diagram shows several clear spectrum peaks and broadband
noise. On the basis of these findings, it was concluded that the observed
fluctuation was caused by several simultaneous process events.

Figure 12.22. Yield spectrum diagram

A MAR analysis was performed separately for several frequency ranges. The
strongest spectrum peaks and broadband noise were included in the analysis.
The proportional fluctuation strength of the strongest spectrum peak measured
more than half the total fluctuation strength, as the Periodic fluctuations dialog
shows.

93
Figure 12.23. Periodic fluctuations dialog

In the fluctuation strength percentage table in the Periodic fluctuations dialog, the
biggest fluctuations in terms of strength were selected, the time scale for which is
shown in the MAR Navigator dialog.

Figure 12.24. MAR Navigator Startup dialog

In the MAR Navigator outcome dialog, you can see that the fluctuation was caused
by the temperature fluctuations of the process steam supplied from outside the
plant.

94
 Figure 12.25. MAR Navigator outcome

The origin of the broadband noise also needed to be clarified. With the MAR
Navigator, the source of the noise was traced to the vicinity of a tank inlet flow.
With the assistance of the plant personnel, the problem was quickly resolved.
Firstly, the tank level was measured with ultrasound. Foam build-up in the tank
interfered with the measurement, however. The resulting fluctuation was observed
at several measurement points throughout the plant.
The sources of several other spectrum peaks was also successfully located. These
peaks had originated from vibrating control circuits.

Non-recurring phenomena
The occurrence of non-recurring phenomena in process measurements can be the
result of, for example, damaged or deteriorated process equipment, equipment
maintenance, modification of control circuit set values or other measures
performed by the operating personnel that affect the process. This type of
phenomenon usually also affects other areas of the process and can often be
observed as a corresponding change in other process measurements.

Waveform matching
The Waveform matching analysis tool searches other measurement signals for
the selected time interval waveform. Waveform matching also finds reverse forms
(mirror images of the waveform).
To begin waveform matching, a single measurement must be selected. Open a
trend window for the selected measurement and, using the X-limiter, select from
the trend plot the phenomenon to be analysed, in other words, the time interval
during which the waveform in question occurs. Right-click on the trend plot and
select either of the following:
• Search for causes of waveform
• Search for consequences of waveform.
If you select the Search for causes of waveform option, the search will be limited
to the time preceding the selected phenomenon. Knowledge of the process is
required for the time interval selection in order to be able to anticipate delays. Too
long a history can cause the function to match the analysed phenomenon with an
event that is so far removed in terms of time that it could never have caused the
phenomenon being investigated. Then the actual cause of the phenomenon in
question can thus remain unfound.

You can limit the search by selecting an already existing measurement group or by
creating a totally new measurement group for the analysis.
The result will be shown in the Waveform matching dialog. The forms that most
closely correspond to the waveform of the selected measurement are presented in
a table. The table shows the name of the measurement, the level of similarity
expressed as a percentage and the delay with which the similarity occurs.

95
Figure 12.26. Waveform matching dialog

Trend plots are also obtained as a result of the analysis. The topmost diagram
shows a measurement selected as the target of the analysis. The time interval in
question is shown on the diagram in a blue colour. The other trend plots present
the other signal forms that best correspond to the target waveform. The identified
waveforms are shown on the trend plots with a yellow background.

96
Figure 12.27. Search for causes of waveform

It is important to note that the Waveform matching function does not take into
account process dynamics. Thus, for example, filtering caused by a tank can result
in the function failing to identify the actual cause of a phenomenon occurring after
the tank.

If you select the Search for causes of waveform function, the search will be
limited to the time after the selected phenomenon. This enables you to search, for
example, the consequences of variation detected in raw material later in the
process.

97
13. Principal Component Analysis (PCA)
This chapter discusses principal component analysis. Principal component analysis
(PCA) is a tool designed for studying a large group of correlated measurements.
PCA enables easy monitoring of process status and the variables that have an
effect on it. A source of disturbance can be located quickly in the measurements
being examined.

In this chapter, you will learn the following:

• Creating PCA models
• Using PCA models
• Deleting PCA models.

Theory of PCA modelling

Industrial processes give a vast number of measurement signals. It is impossible to
monitor all signals continuously. However, various process variables have an effect
on each other, in other words, they correlate. The data contained within these
correlated signals can be expressed using a smaller number of calculated variables
than original variables. Principal component analysis (PCA) can be used to pack
this data. The aim, therefore, is to express the information contained in the original
variables using the new variables while retaining as much information as possible,
even though the number of the new variables is much lower than that of the
original ones. New variables are formed by forming linear combinations of the
original variables. This means multiplying each original variable by an appropriate
factor and adding up the multiplied variables.

The fluctuations of two correlating variables can be expressed reasonably well

using a single new combined variable Y, which is a so-called principal component.

Y i = P1  X1 + P 2  X 2

The factors Pi express the degree of influence each factor xi has on the values of
the combined variables Yi.

When creating a PCA model, the data is used to determine the directions in which
fluctuation most occurs. These directions are the new coordinate axes (principal
components), and their directions remain unchanged after the model has been
created. These new coordinate axes are used to express the fluctuation in original
measurement signals. Principal components do not correlate with each other.

98
Creating PCA models
Steps for successful PCA analysis:
1. Choose interesting measurements.
• no long process delays
2. Choose appropriate periods of time for comparison data.
• process in a typical state
• good product quality
• time periods can be cut
• no constant signals
3. Create a PCA model.
4. Study the results.

To create a PCA model:

1. Choose a reference period to base the PCA model on.

• All the time period selection properties of Wedge can be used when selecting
the reference period. Data cutoff and screening can also be used. State
dependency should be observed also when creating a PCA model. If state
dependency is used for defining limits, it is recommended that several periods
during which the process has been in the same mode are included in the
reference period.
• If you want to save the cut time interval, save it as a Favorite time interval.
2. Select the measurements that you want to create a PCA model for. If no
measurements are selected, a PCA model is created using all measurements in
the measurement group.

3. Select PCA Model from the Analyses menu. The PCA dialog will appear in the
main window on the right.

Figure 13.1. PCA dialog

4. Select Create PCA model. The Create PCA Model dialog will appear.

5. Select whether the PCA model will be created for the measurements of the
selected measurement group or the selected measurements and click Next.
6. Enter a name and a confidence limit for the PCA model. Select an appropriate
state dependency option. Click Next.

99
 7. Enter the number of principal components to be calculated. The dialog displays
the influence that the number of principle components has on the percentage of
explained variance. The red colour in the bar chart signifies the proportion of
fluctuation attributable to the selected principal components. The default number of
principal components is three. This number can be changed, upon which the
cumulative percentage of explained variance also changes.

There is no universal rule for the number of principal components selected for
analysis or for an adequate cumulative percentage of explained variance.
However, the cumulative explained variance of the principal components selected
for analysis should be more than 50% for the process data analysed.

8. Select Finish.

Figure 13.2. Percentage of explained variance of principal components

Once the selections have been approved, Wedge calculates the selected number
of principal components. The principal component icons contain the calculated
data of the principal components. In addition, the PCA group includes one icon for
residual variance. The principal component and residual variance icons include the
same functionality as the normal measurements.

Figure 13.3. Principal component icons

The PCA model is saved along with the process diagram and can be reapplied.

Using PCA models

Principal components differ from other measurements in the following ways:

100
 • They cannot be deleted.
• They cannot be moved or copied to other measurement groups.
A PCA model can also be state-dependent.

The results of principal component analysis consist of calculated variables (i.e.,

principal components) expressed as time series and confidence limits that indicate
their normal variation. Principal components can be used, for example, to observe
whether the process was in the same state at the target time as it was during the
reference period. If the principal component values do not cross their confidence
limits, it can be assumed that as far as the measurements included in the PCA
model are concerned, the process is in the same state as it was during the
reference period. If any principal component crosses its confidence limit, the PCA
can be used to go back to the original signals to find out what values show the
most significant changes compared to the reference period.
PCA also helps to identify situations where the correlation of measurements has
changed. This means that there has been a significant change in the normal
correlation of certain measurement signals. Typically, this refers to incorrect values
received from sensors or an exceptional process state.

Studying principal components individually

When a confidence limit of a principal component has been crossed, you can easily
study the original measurements used to create the PCA model. Thus, you can find
out why the value crossed the limit.

Do as follows:

1. Select a point of time from the principal component trend for further examination.

2. Right-click on the trend plot.

3. From the menu, select Contributions.

• The program displays a bar chart with as many bars (factors) as there were
original measurements in the PCA model. The bar chart describes the effect of
each original variable on the principal component value at the selected point of
time. When the values of the bars are added up, the result is the value of the
principal component at the selected point of time. A high bar value refers to a
high impact on the principal component value at the selected point of time. The
diagram usually indicates clearly which variable has caused the crossing of
confidence limits.

101
Figure 13.4. Principal component contributions

4. Use the tooltip function to see which variable a bar refers to. To show the tooltip,
move the mouse pointer onto the bar.

Alternatively, you can do as follows:

1. Select the desired point of time from the principal component trend plot.

2. Right-click on the trend plot.

3. Select Top4. A trend plot is displayed with the examined principal component at
the top. Below the principal component, you can see the trends of the four original
measurements that best indicate the value of the selected point of time.

102
Figure 13.5. Principal components Top4

To view the differences between two points of time:

1. Select two points of time in the principal component trend.

2. Right-click on the trend plot.

3. From the menu, select Difference contributions. A bar diagram is displayed.

The bar values represent the factors that are responsible for the difference in the
principal component values between two points of time. The bars describe the
change in the original measurements between the selected points of time.

Alternatively, you can do as follows:

1. Select two points of time in the principal component trend.

2. Right-click on the trend plot.

3. Select Difference Top4. A window is displayed with the examined principal

component at the top. Below the principal component, you can see the trends of
the four original measurements that have changed the most between the selected
points of time.

103
Updating principal component models
To update a PCA model:

1. Select a point of time on the basis of which the model will be updated.

2. In the PCA dialog, select the Update PCA model.

3. You can also change the name and confidence limit of the PCA model. If you
want to make further changes to the PCA model, create a new model.

4. Click Update.

Effect of original variables on principal components

Use the Loadings... function in the PCA dialog to view which original
measurements have the most effect on each principal component.

Do as follows:

1. Select the principal components whose structure you want to study.

2. Select Loadings.... A Trend window will appear, showing the effect of the
original measurements on various principal components as bar diagrams.

Figure 13.6. Loadings dialog

The bars of the diagram indicate the direction and amount of the effect that the
original variables have on the principal component values. The original signals
whose bars are the largest (in either the positive or the negative direction) have the
most effect on the principal component in question. A positive bar value indicates a
positive correlation between the measurement signal and the principal component.
A negative value indicates a negative correlation.

104
XY score (effective score)
The XY score diagram shows the time series of two selected principal components.
The diagram shows the behaviour of the process during the examination period in
regard to the selected principal components. If points form clear groups in the
diagram, the process has been in various states during the period. These process
states can be compared by viewing the XY score and XY loadings diagrams
simultaneously. However, it is recommended that a separate principal component
analysis be performed for each process state and that the results of these analyses
be then compared to each other.

The points should be inside the ellipse. If the points are clearly outside the ellipse,
there has been a change in the process since the time period used when creating
the PCA model.

Figure 13.7. XY score

105
XY loadings
The XY loadings diagram shows the interdependencies of the variables included in
the PCA model. The following rules are followed when studying the
interdependencies of the variables:
1. The closer to the origin the variable is, the less weight it has in the creation of
the principal components.

2. Correspondingly, the further from the origin a variable is, the more important it is.
3. Variables that are close to each other have a positive correlation.

4. Variables that are located on the opposite sides of the origin have a negative
correlation.
5. The closer the angle of the straight lines drawn from two variables to the origin is
to a right angle, the weaker the correlation of these variables.

Figure 13.8. XY loadings

Use the XY loadings diagram to, for example, find out how the correlations
between certain variables differ when different states are considered.

106
PCA model residual variance
When the PCA model was created, an icon for residual variance was created in the
measurement diagram in addition to principal component icons. Residual variance
includes the variance that is not explained by the principal components of the
model. The same functions can be performed for residual variance as for the
principal components. The residual variance diagram shows clearly the points
where there are the most deviations in the correlations between the variables.
When a deviation occurs, the residual variance trend crosses the confidence limit.

Figure 13.9. Residual variance

Deleting PCA models

To delete a PCA model:

1. Select PCA Model from the Analyses menu.

2. In the PCA dialog, select Delete PCA model.

3. Click Yes.

4. Save the diagram.

107
14. Macros
This chapter deals with creating macros. A macro includes a series of frequently
used Wedge functions that can be performed by a single command.

In this chapter, you will learn the following:

• Recording macros
• Executing macros
• Saving macros
• Editing macros
• Deleting macros
• Create background macros.

Recording macros
Since saved functions can be repeated any time and as often as necessary,
creating (recording) macros saves time and makes Wedge easier to use.

To record a macro:

1. From the Tools menu, select the Create Macro... command. The Macros dialog
will appear.

2. Enter a name for the new macro in the New Macro text box.

Figure 14.1. Macros dialog

3. Click the Record… button. The dialog will close and the text REC will appear on
the status bar at the bottom of the main window.
4. Include the functions in the macro. The following is an example of a function
series:
• Select Last day as the time interval.
• Select the measurements of your choice.
• Open the trends of the measurements.
• Open the statistics of the measurements.
5. When you have recorded the function series, select the End Macro command
from the Tools menu.

108
Executing macros
To execute a macro, select the Run macro command from the Tool menu and
select the macro.

Executing macros via a diagram macrolink

The quickest way to use macros in a process diagram is to add a macrolink.
Adding a macrolink component to a diagram is presented in the chapter Editing a
diagram.
To execute a macro, simply click the macrolink.

Saving macros
Recorded macros are saved automatically when saving the process diagram.
When the process diagram is read from the file, the macros that existed when
saving the diagram will be available again.

Editing macros
You can edit a recorded macro using WedgeLISP commands. WedgeLISP closely
resembles the LISP programming language.

To edit a macro you have created:

1. From the Tools menu, select the Create Macro... command. The Macros dialog
will appear.

2. In the All Macros box, select the macro you wish to edit. Click the Edit… button.
The Edit Macro dialog will appear. Add LISP commands or edit and delete existing
commands in the text box.

109
 Figure 14.2. Edit Macro dialog

3. When you have made all the necessary changes in the Edit Macro dialog, click
Apply and Close. The Edit Macro dialog will close.

4. Click the Close button in the Macros dialog.

Deleting macros
To delete a macro, select it in the All Macros list box in the Macros dialog and
click Delete. If a process diagram is saved after deleting a macro, the macro will no
longer be available.

Background macros
Background macros can be used to perform certain tasks automatically at specific
time intervals without disturbing the normal use of the Wedge system.

One function of background macros is to report the results of a given function

(measurement, analysis, etc.) at regular intervals as a trend plot. The diagram
related to the function must be selected in the process display window.

110
Setting a macro as a background macro
To set a macro as a background macro:

1. From the Tools menu, select the Create Macro... command. The Macros dialog
will appear.
2. Select a macro from the list in the Macros dialog and click the Tasks... button.
The Background macros dialog will appear.

Figure 14.3. Background macros dialog

3. Enter the time interval between consecutive macro executions in the Task
Interval text box (for example, 30). The minimum time interval is five seconds.

4. Select Select the macro to be used as a background macro from the Macro List
list box. To move the macro into the Task List list box, click the upper arrow
button.

5. Click Close.

Interrupting background macro execution

To delete a macro from the background macro list:

1. From the Tools menu, select the Create Macro... command. The Macros dialog
will appear.

2. Click the Background… button. The Background macros dialog will appear.

3. Click a macro name in the Task List list box.

4. Click the lower arrow button. The macro is moved back to the Macro list list box.

5. Click Close.

111
15. Printing
This chapter discusses printing. You can plan a layout for a page in the Page
Composition dialog. Use report templates and drawing tools to assist in the
layout. If a dialog has a Print… button, a page layout does not have to be planned
separately. If you want to print, for example, the trend plots displayed in the Trend
window, just click the Print… button and define the print settings in the Printer
Settings dialog.

Page composition planning

Use the Page Composition dialog to plan the layout. You can find the Page
Composition command in Wedge's File menu.

Figure 15.1. Page Composition dialog

112
Adding basic report templates
Use the Add Report Page list box at the top of the print dialog to select a report
template. To add a basic report template to the page:

1. Click the drop-down arrow of the Add Report Page list box.

2. Select a basic report from the list. The Report Page dialog will appear.

Figure 15.2. Basic Report Page dialog

3. You can add several details to the report. If you want to:
• add a header, select the Add Header check box and enter the header text in the
text box.
• add a time interval, select the Add interval check box.
• add a sampling interval, select the Add time level check box.
• add either statistical indicators or trends only or both together, select the Show
on report page check box.
• set the maximum amount of trends per column by entering a value in the check
box (default is 8).
• select the number of columns in which trends are to be drawn by making the
selection in the check box (default is two columns).
4. Click OK. The selected report template is added to the page.

113
 Figure 15.3. Completed basic report

Adding table report templates

Use the Add Report Page list box at the top of the print dialog to select a report
template. To add a table report template to the page:
1. Click the drop-down arrow of the Add Report Page list box.

2. Select a table report from the list. The Report Page dialog will appear.

114
Figure 15.4. Table Report Page dialog

3. You can add several details to the report. If you want to:
• add a header, select the Add Header check box and enter the header text in the
text box.
• add a time interval, select the Add interval check box.
• add a sampling interval, select the Add time level check box.
• select the basis for the collection, i.e., the type of data that generates new
columns in the table, the options are:
o Day
o Grade
o Optional measurement.
• set the widths of table cells.
• add the total sums of measurements in the report.
4. Click OK. The selected report template is added to the page.

115
 Figure 15.5. Completed table report

Adding printables
You can build a printing page using the contents of open dialogs and diagram, i.e.,
all material displayed. The Add Print Object list box at the top of the dialog shows
all objects that you can add to the page. You can add, for example:
• Diagram images
• Trend plots
• Analysis results
• Time intervals used.

Page Composition dialog tools

Use the drawing tools on the left-hand edge of the Page Composition dialog to
edit the page layout. Move the pointer on top of a button and wait a moment. A
tooltip will appear. Tooltips contain short descriptions of the button functions.
Select the Page Setup... button at the bottom of the Page Composition dialog to
open the Page Setup dialog. As a default, the page is printed on A4 paper in
portrait orientation. The footer can include the page number and/or date and
printing time.

116
Figure 15.6. Print page size settings

117
16. Reports
This chapter deals with creating and printing reports. Often used standard reports
are usually created and printed using the report tool.

In this chapter, you will learn the following:

• Creating standard reports using the report tool
• Printing reports.

Using the report tool

You can create reports with the Report Wizard. The report tool is accessible for use
by the software administrator or reporting administrator. To create a new
automatically printed report:
1. From the Edit menu, select the Settings command and then select the
Reporting option.

Figure 16.1. Reports dialog

2. Select Create new report. Enter a name for the report and select its type. You
can choose a Basic report, a Table report or a report based on a predefined
macro. Once you have named the report, click Next.

118
Figure 16.2. Report naming dialog

3. Next, a dialog will appear in which the layout of the Print page can be selected.
Make the necessary selections: paper size, orientation and footer fields.

Click Next or return to the pervious stage by clicking Back.

Figure 16.3. Report's Print page settings dialog

4. Next, specify the contents of the report. For a basic report, for example, you can
choose whether to display Statistics or Trend plots or both in the report.

119
Figure 16.4. Report contents dialog

5. Next, select the diagram measurement groups for which you wish to report
measurements. Multiple measurement groups can be selected. To remove an
already selected group, select its name and click the Delete group button. You can
change the order of the measurement groups selected for reports using the Up and
Down buttons.

All measurement group measurements are selected for the report.

Click the Group properties button to open the Measurement group properties
window. You can determine the order in which the measurement group trends are
printed on the report pages.

120
Figure 16.5. Report Measurement group selection dialog

6. In the dialog below, you can define the time interval and time level of the
measurement data.

Select the end time of the interval from the following options in the drop-down
menu: Report date, Report time, Weekday or Any given day of the month
1...31.

If, as the end time of the interval, you choose the Report date, a weekday or any
given day of the month, the end time can be set in the time column: hh:dd:mm.
Select the Time Level from the drop-down menu. Make sure that the
measurements selected for the report correspond to the selected time levels and
they have measurement data.

Figure 16.6. Report Measurement time period and time level dialog

7. A report can be scheduled to be drawn up either daily or on a certain day of the

week or month.

Set the execution time for the report with the Add new time command. Select from
the dialog that opens up the printing time in the form hh:mm:ss.
The printing time selection can also be skipped, in which case the report is only
printed manually as necessary. In this case, select the Not repeated check box.
Click Next or return to the pervious stage by clicking Back.

121
Figure 16.7. Report Printing time dialog

8. Finally, assign a printer to the report.

If output to a PDF file is selected, the report will be sent via the Wedge server to a
configured directory. Each report can be named successively according to its
printing/output time and/or date.

Other options include sequential numbering or weekday-specific naming. The

remaining alternative is to print the report in PDF format over the previous report
without changing the name.

122
Figure 16.8. Report printer selection dialog

Once the report is ready, you can use the buttons in the report tool dialog to
manage existing reports.

Figure 16.9. Completed report

To edit an existing report, select the report from the list and click the Modify report
button.
To copy an existing report under a different name, select the report from the list
and click the Copy report button. The report, complete with settings, is copied

123
 under a new name.

To delete an existing report, select the report from the list and click the Delete
report button.

To see a Report preview, first select the report to be previewed from the list. Then
click Preview. Wedge retrieves the measurement signals and measurement data
defined for the report and opens the Page Composition dialog. You can browse
the print pages to check the final print layout of the report.

Report time intervals

To define a separate macro for use in a report macro, you can select the time
interval using the separate Report time interval tool.
The time interval refers to the start and end times of the data to be processed.
Define the time range before creating a report. Wedge will then process the correct
data.

To define a time interval for a report

1. From the Tools menu, select the Report Time Interval... command. The
Report Time Interval dialog will appear.

Figure 16.10. Report Time Interval dialog

2. Define the length of the time interval in the combo box on the left by setting the
following:
• Number of months
• Number of days
• Time.
3. To select the end point of the time interval:
• If the Use interval end time check box is selected, you can select the time
interval end date in the list box at the centre of the dialog and a time in the
combo box on the right.
• If the Use interval end time check box is empty, Wedge will set the end point
of the interval as the report execution time.
3. Click OK to close the dialog.

Scaling of trend plot Y-axes in reports

You can define the Y-axis scaling to be used in reports with the Trend scaling...
tool. To set fixed Y-axis scaling:

124
1. Select the measurements for which you want to set fixed Y-axis scaling.

2. Choose Trend scaling... in the Options menu.

3. Select the measurement for which you want to fix the Y-axis scaling of the trend
plot. Click the Minimum field in the column: enter the minimum Y scale. Select the
corresponding Maximum field in the column and enter the maximum value for the
Y-axis.

4. Do the same for all measurements for which you want to fix the trend plot scaling
in Reports.

4. Once you have set the maximum and minimum values, click OK.

5. Save the diagram. The Y-axis scalings of the trend plots are saved along with
the diagram file.

Figure 16.11. Trend Y-axis scaling in Reports dialog

125
17. Saving measurement data
This chapter deals with saving Wedge measurement data as a text file in ASCII
format. You can use ASCII saving if you want to process Wedge data in another
program. Measurement data can also be moved directly to Excel via the clipboard.
In this chapter, you will learn the following:
• Saving measurement group data as text file
• Transferring measurement data to Microsoft Excel via the clipboard.

Saving measurement data

You can save measurement data in ASCII format only when the process diagram is
open. If no measurements are selected, Wedge saves data from all measurements
by default.

Figure 17.1. Save measurement data as ASCII File dialog

• The Save to ASCII file... command in the Tools menu opens the Save to
ASCII file dialog.
The dialog contains the following options:
• Save in DAS format
• Time format
• Column separator
• Decimal symbol
• File name

126
 Column delimiter is used to separate the measurement data of various
measurements made at different times, when this data is written on one line. The
default name is [Link] and the default location is the user's desktop.

The DAS format option edits the file to the default format.

To save measurement data to an ASCII file:

1. Select the measurement(s) from the process diagram and ensure that the
measurement contains the necessary data.
2. Choose Save to ASCII file... in the Tools menu. The Save to ASCII File dialog
will appear.

3. Select the time format, column delimiter and decimal symbol.

4. Click the button in the lower right-hand corner to open the Enter File Name
dialog. Use the dialog to define the directory for the file and the file name. Click
Save.

Copying measurement data onto the clipboard

You can copy and transfer measurement data directly to Microsoft Excel via the
clipboard.

To save the data from one or several measurements via the clipboard:

1. Select the measurements from the process diagram.

2. From the Edit menu, select the Copy data to the clipboard command.

3. Start Microsoft Excel and open a new workbook. Paste the measurement data
into the workbook using the Paste command (or the Ctrl-V keyboard shortcut).

127
18. Importing data from a worksheet
This chapter deals with importing data from an Excel worksheet to be used in
Wedge. Importing data requires administrator rights and that Excel be installed in
the user's computer.
The imported data will be saved in an SQL database stored in the Wedge server
and is available for use by all users. This data import tool is designed for interactive
import of small amounts of data, such as for importing test run results into Wedge.

Data import
The Import data from a worksheet command in the Tools menu opens the
Worksheet data dialog.

Figure 18.1. Worksheet data dialog

In this dialog, you can choose whether to import a new worksheet or to make
changes to earlier imports. This chapter presents first the import of new
worksheets. Select the "Import new worksheet file" link, which opens the following
dialog:

128
Figure 18.2. Worksheet selection

Press the button in the upper right-hand corner that is circled in the figure. This
opens a file listing where you can select the worksheet of your choice.

Sheet: drop-down menu, you can select the Excel tab where the data to be
imported is located.

The found data is shown in the Data preview dialog.

Figure 18.3 Data preview dialog

Click the Next button to open the Cell settings dialog, where you can specify the
first cell containing the names of the measurements. In the same dialog, you can
activate the Unit and Label cells and specify their locations.

129
Here you can also change the direction of measurement and time fields
(horizontal/vertical).

Data import supports worksheets where dates and times are in one or two columns
(or rows).

Figure 18.4. Cell settings dialog

Click the Advanced button to open the Advanced options dialog, where you can
specify more detailed settings.

130
 Figure 18.5. Advanced options dialog

In this view, you can choose to specify that the name of the measurement will be
collected from several columns or that the size of the worksheet to be imported will
be limited.

Click the Next button to open the Data source selection dialog, where you can
choose whether a new data source will be created for the data or whether the data
will be added to an existing data source. Finally, click Finish to start the data
import.

Figure 18.6. Data source selection dialog

Management of imported data sources

The Worksheet data dialog also contains the Manage spreadsheet data sources
option. This opens the Data source selection dialog. Use the topmost drop-down
menu Data source: in the dialog to select the data source that you want to edit.

In the Imported from: drop-down menu, you can see the files imported to the data
source in question. Press the Delete import button at the bottom of the dialog to
delete the file data from the data source. Press the Delete source button at the

131
bottom of the dialog to delete the selected data source completely.

Figure 18.7. Data source selection dialog

132
19. User-specific settings
This chapter deals with defining user-specific program settings. You can find the
Settings command in Wedge’s Edit menu. When you select the Settings
command, the Settings dialog will appear, on the left-hand side of which is the
Directory option. This chapter discusses the settings for the following functions:
• Process diagrams
• Time intervals
• Time settings
• Main window settings
• Servers.

Process Diagrams
Use the Process Diagrams view to select the process diagram default directory
and favorite process diagrams. Set the most commonly used directory as the
process diagram default directory. Then set the most commonly used process
diagrams as the favorite process diagrams.

Figure 19.1. Process Diagrams view of the Settings dialog

Default Process Diagram Directory

To define a process diagram default directory:
1. Click the button in the right-hand corner of the Default Process Diagram
Directory field. The Select Directory Name dialog will appear.

133
 2. Click the drop-down arrow of the Directory field, select the directory from the list
and click OK. The Select Directory Name dialog will close and the selected
default directory will appear in the Default Process Diagram Directory field.

3. Click Save.

Favorite Process Diagrams

Favorite process diagrams refer to often used diagrams that the user has set as
favorites. Favorite diagrams speed up the use of the program and their names
appear in the Open Favorite submenu of the File menu.

To add a new favorite process diagram:

1. Click the button in the Favorite Process Diagram section at the bottom of the
dialog. The Select File Name dialog will appear.

2. Select a folder and a diagram.

3. Click Save. The dialog will close.

4. Click the Add button in the Settings dialog. The selected file or folder will
appear in the Favorite Process Diagrams list.

5. Click Save and then Close.

To change the order of diagrams in the favorites list:

• Click the name of the process diagram in the favorites list.
• Move the diagram up or down using the arrow buttons in the lower right-hand
corner.

To remove a process diagram from the favorites list:

• Click the name of the process diagram in the favorites list.
• Select Delete, Save and Close.

134
Time intervals
In the Time intervals view, you can edit user-specific time interval settings.
In this section, you will learn the following:
• Setting time interval shortcuts
• Setting favorite time intervals
• Setting default intervals
• Starting the program in the on-line display mode.

Figure 19.2. Time Interval Shortcuts tab of the Settings dialog

Time Interval Shortcuts

To define time interval shortcuts in the Time Interval list box:
1. Enter a name for the shortcut in the text box at the bottom of the tab.

2. Update the number of days, the time or both in the combo box.

3. Select Add, Save and Close.

To change the order of time interval shortcuts:

• Click the time interval name in the shortcut list.
• Move the interval up or down using the arrow buttons in the lower right-hand
corner.

To remove a time interval from the shortcuts:

• Click the name of the interval in the favorites list.

135
 • Select Delete, Save and Close.

Favorite Time Intervals

Use this function to add the current time interval to the favorite time intervals (if the
time interval contains data that you want easy access to for later processing).
To add a new favorite time interval:

1. Ensure that the measurements contain the necessary data.

• You can cut time intervals that contain faulty data, such as process disturbance
periods, from the data. The cut sections of the time axis of the data are saved
along with the favorite time interval data.
2. Check that the time interval is displayed in the Set Time Intervals list box on the
shortcut bar.

3. Click the Favorite Time Intervals tab on the right.

Figure 19.3. Favorite Time Intervals tab of the Settings dialog

4. Enter the favorite time interval name in the Name field.

5. Click the Add current interval button. The time interval name will appear in the
Favorite Time Intervals field.

6. To check the content of a saved interval, click its name in the list. The interval
will appear at the bottom of the Favorite Time Intervals field.

7. Click Save and then Close.

To change the order of favorite time intervals:

• Click the favorite time interval name in the shortcut list.
• Move the interval up or down using the arrow buttons in the lower right-hand
corner.

136
 To remove a time interval from the favorites:
• Select the interval name in the list.
• Select Delete, Save and Close.

Default level and default interval

A default level refers to the time level that appears as the default in the opened
software shortcut bar. A default interval refers to the interval that appears as the
default in the opened software shortcut bar.

To select a default level and default interval:

1. Click the Default interval tab on the right.

2. Select a time level from the Default Level list box.

3. Select a time interval from the Default Interval list box.

4. Click Save and then Close.

Figure 19.4. Default Interval tab of the Settings dialog

On-line display
The time interval shown by Wedge can also be dynamic, i.e., it is updated as time
progresses. This means that the opened Trend window is updated as new process
data accumulates. The update speed of the Trend window depends on the
selected time level. To set the dynamic trend display program as the default, click
the Start program in on-line display mode check box.

137
Time Settings
The time settings dialog enables you to set the date and time format.

Figure 19.5. Time Settings view of the Settings dialog

The default format for the date is [Link], where DD refers to days (01–31),
MM to months (01–12) and YYYY to years (0000–9999).

To change the date format:

1. Select a date format from the Date Format list box. The following five options
appear in the list:
• [Link] (default)
• DD/MM/YYYY
• MM/DD/YYYY
• YYYY-MM-DD
• MM DD YYYY.
2. Click the preferred format. The selected format will appear in the selection field.

3. Click Apply and then Close.

To change the time format:

1. Choose a time format from the Time Format list box. The list contains the
following options:

- HH:MM:SS (default)
- [Link].

2. Click the preferred format. The selected format will appear in the selection field.

138
3. Select Apply and then Close.

Wedge contains the following time formats:

Option Description

The 24-hour clock is used.

If this option is not selected, the 12-hour clock
24-Hour Clock is used.

Monday is the first day of the week.

If this option is not selected, Sunday is the first
Week Starts day of the week.
Monday

The names of the months will be shown.

If this option is not selected, the names of
Show Months months will not be shown.

The day of the week will be shown.

If this option is not selected, days of the week
Show Weekdays will not be shown.

To enable one of the formats mentioned above:

• Click the check box to add a selection mark.
• Click Apply and then Close.

139
Main Window Settings
In the Main Window Settings view of the Settings dialog, you can edit the
following:
• Measurements lists
• Measurement captions
• Tooltip instructions.

Figure 19.6. Main Window Settings view of the Settings dialog

Measurements list settings

In the Measurements list settings, you can select which columns are shown in the
measurements list. The options are:
• Name
• Unit
• Description
• Measurement data source position
• Data source.

Setting a measurement component caption

Use the Measurement Caption list box to select one of the following options as
the measurement icon caption that is displayed in the diagram:
• Measurement name
• Description
• Measurement tag name in the data source.

140
Tooltip instructions settings
In the Tooltip instructions field, you can determine whether tooltip instructions for
measurements are shown on the canvas. There are two options to choose from:
• Show tooltips always.
• Show tooltips when editing diagrams.
Tooltip instructions appear automatically when the mouse pointer is placed over a
measurement component. The tooltip instruction gives information on the label of
the measurement and the data source name and position in the data source.

141
Servers
When you start Wedge, it automatically connects to the servers that were
predefined during installation. The servers in use are specified in the Servers view
of the Settings dialog that is opened via the Edit menu. Whenever necessary, you
can define a new server or restore the default settings. Only administrators can edit
server settings, however.

Figure 19.7. Servers view of the Settings dialog

WCS (Wedge Configuration Server) is a configuration server that is used to save,

for example, user-specific settings (such as startup settings) and data source
channel information.

The data server DAS (Data Acquisition Server) is a software that Wedge uses to
read measurement data from various data sources. Only one data server can be
activated in a server computer at once. Data server have at least one data source
per each data collection device connected to them. Each data source normally has
at least one measurement.

The alarm server program WES (Wedge Event Server) is used by Wedge to
monitor the status of the automatic process monitoring system. The alarm server
transmits alarms to the client programs that are connected to them.

The file server program WFS (Wedge File Server) is used by Wedge to save and
archive diagram files on the server.
The MATLAB server enables server-based performance of calculations. At startup,
Wedge connects automatically to a MATLAB server predefined during system
installation.

142
To define a new server:

1. From the Settings dialog directory, select the Servers view.

2. Click Set current.

3. Enter the server name or IP address in the dialog that opens up.

4. Click OK to accept the settings.

Use the Default value button to set a server that is in use as default during the
next Wedge startup.

143