AUTOMATED CONTROLS

MEN 435
Lecture 9

Stability

Automated Controls
MEN 435 Sheet 1
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction

Automated Controls
MEN 435 Sheet 2
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
Stability is Most Important System Specification
• If System is Unstable
• Transient Response and
• Steady State Errors
are Moot (Unsettled) Points
• Unstable System cannot be designed for
– Specific Transient Response
or for
– Steady State Error Requirement
• Output of System could only be controlled if
– Steady-State Response consists of only Forced Response
Automated Controls
MEN 435 Sheet 3
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• But Total Response of System is
– Sum of Forced and Natural Responses
c(t) = cforced(t) + cnatural(t) (6.1)

• Using these Concepts, given in the Next Slides the Definitions of

• Stability
• Instability and
• Marginal Stability

Automated Controls
MEN 435 Sheet 4
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Linear, Time-Invariant System is
• Stable if
• Natural Response approaches Zero as
• Time approaches Infinity
• Unstable if
• Natural Response grows without Bound as
• Time Approaches Infinity

Automated Controls
MEN 435 Sheet 5
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Marginally Stable if
• Natural Response
• neither decays
• nor grows
but
• remains Constant or
• oscillates
as Time approaches Infinity
• Thus
• Stability implies that only
• Forced Response remains
• as Natural Response approaches Zero
Automated Controls
MEN 435 Sheet 6
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• These Definitions rely on Description of Natural Response
• When looking at Total Response
• it may be difficult to separate
• Natural Response from Forced Response
• However, we realize that
• if
• Input bounded and
• Total Response not approaching Infinity as
• Time approaches Infinity
• then
• Natural Response obviously not approaching Infinity
Automated Controls
MEN 435 Sheet 7
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• If Input unbounded
• we will see Unbounded Total Response, and
• we cannot arrive at any Conclusion about Stability of System
• We cannot tell whether Total Response is unbounded because
• Forced Response unbounded or
• Natural Response is unbounded

Automated Controls
MEN 435 Sheet 8
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Thus, Alternate Definition of Stability one that
• regards Total Response and
• implies 1st Definition based upon Natural Response
is this:
System is Stable if
• Every Bounded Input yields Bounded Output
• We call this Definition of Stability
Bounded-Input, Bounded-Output (BIBO)

Automated Controls
MEN 435 Sheet 9
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Let us now produce
• Alternate Definition for Instability
• based on Total Response
• rather than Natural Response
• We realize that
• if Input Bounded but Total Response unbounded
• System Unstable
• since we can conclude that
• Natural Response approaches Infinity as
• Time approaches Infinity

Automated Controls
MEN 435 Sheet 10
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• If Input Unbounded
• we will see Unbounded Total Response and
• we cannot draw any Conclusion about Stability of System
• we cannot tell whether Total Response is Unbounded because
• Forced Response is Unbounded or
• Natural Response is Unbounded
• Thus, Alternate Definition of Instability,
• One that regards Total Response, is this:
• System Unstable
• if any Bounded Input yields Unbounded Output

Automated Controls
MEN 435 Sheet 11
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• These Definitions help clarify our previous Definition of
Marginal Stability
which really means that System is
• Stable for some Bounded Inputs and
• Unstable for Others

Automated Controls
MEN 435 Sheet 12
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• For Example, we will show that
• if Natural Response Undamped
• Bounded Sinusoidal Input of same Frequency as Natural
Response yields
• Natural Response of growing Oscillations
• Hence System
• appears Stable for all Bounded Inputs
• except this One Sinusoid
• Thus, Marginally Stable Systems by Natural Response Definitions are
included as
• Unstable Systems under BIBO Definitions
Automated Controls
MEN 435 Sheet 13
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Summarizing Definitions of Stability for Linear, Time-Invariant Systems
• Using Natural Response
1. System is Stable if
• Natural Response approaches Zero as Time approaches Infinity
2. System is Unstable if
• Natural Response approaches Infinity as Time approaches
Infinity
3. System is Marginally Stable if
• Natural Response neither decays nor grows but remains constant
or oscillates
Automated Controls
MEN 435 Sheet 14
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Using Total Response (BIBO)
1. System is Stable if
• every Bounded Input yields Bounded Output
2. System is Unstable if
• any Bounded Input yields Unbounded Output

Automated Controls
MEN 435 Sheet 15
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Physically, Unstable System whose Natural Response grows without
Bound can cause Damage to
• System
• Adjacent Property or
• Human Life
• Many Times
• Systems are designed with Limited Stops to prevent Total Runaway
• From Perspective of Time Response Plot of Physical System
• Instability is displayed by Transients that grow without Bound
and consequently Total Response that does not approach
• Steady-State Value or other
• Forced Response
Automated Controls
MEN 435 Sheet 16
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• How to determine if System is Stable ?

• Let us focus on Natural Response Definitions of Stability

• Recall from our Study of System Poles that

• Poles in Left Half-Plane (lhp) yield

• either Pure Exponential Decay

• or Damped Sinusoidal Natural Responses

Automated Controls
MEN 435 Sheet 17
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• These Natural Responses decay to Zero as Time approaches Infinity
• Thus, if Closed-Loop System Poles are in
• Left Half Plane
and hence
• have Negative Real Part
System Stable
• That is
• Stable Systems have Closed-Loop Transfer Functions with
Poles only in Left Half-Plane

Automated Controls
MEN 435 Sheet 18
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Poles in Right Half-Plane (rhp) yield
• either Pure Exponentially Increasing
• or Exponentially Increasing Sinusoidal
Natural Responses
• These Natural Responses approach Infinity as Time approaches
Infinity
• Thus, if Closed-Loop System Poles
• are in Right Half of Plane
and hence
• have Positive Real Part
System Unstable
Automated Controls
MEN 435 Sheet 19
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Also, Poles of Multiplicity greater than 1 on Imaginary Axis lead to
Sum of Responses of the Form
A tn cos( t+ )
where n = 1,2,...,
where Amplitude approaches Infinity as Time approaches Infinity
• Thus
• Unstable Systems have
• Closed Loop Transfer Functions with at least
• One Pole in Right Half-Plane and/or
• Poles of Multiplicity greater than 1 on Imaginary Axis

Automated Controls
MEN 435 Sheet 20
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Finally, System that has Imaginary Axis Poles of Multiplicity 1 yields Pure
Sinusoidal Oscillations as Natural Response
• These Responses
• neither increase
• nor decrease
in Amplitude
• Thus
• Marginally Stable Systems have Closed-Loop Transfer Functions with
only
• Imaginary Axis Poles of Multiplicity 1 and
• Poles in Left Half-Plane
Automated Controls
MEN 435 Sheet 21
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• As Example
• Unit Step Response of Stable
System of Figure (a) is compared
to that of Untsable System of
Figure (b)
• The Responses, shown in Figure,
show that while Oscillations for
Stable System diminish, those for
Unstable System increase without
bound
• Also notice that Stable System’s
Response in this Case approaches
Steady-State Value of Unity
Automated Controls
MEN 435 Sheet 22
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• It is not always Simple Matter to determine if
• Feedback Control System is Stable
• Unfortunately, Typical Problem that arises is shown in the Figure
• Although we know Poles of Forward Transfer Function in Figure (a)
• we do not know Location of Poles of Equivalent Closed-Loop System
of Figure (b) without factoring or otherwise solving for Roots

Automated Controls
MEN 435 Sheet 23
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• However, under certain Conditions
• we can draw some Conclusions about Stability of System
• First
• if Closed-Loop Transfer Function has only Left-Half-Plane Poles
• then Factors of Denominator of Closed-Loop System Transfer
Function consist of Products of Terms such as (s+ai)
• where ai is
• Real and Positive or
• Complex with Positive Real Part
• The Product of such Terms is
• Polynomial with all Positive Coefficients
Automated Controls
MEN 435 Sheet 24
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• No Term of Polynomial can be missing
• since that would imply Cancellation between
• Positive and Negative Coefficients or
• Imaginary Axis Roots in Factors
which is not the Case
• Thus, Sufficient Condition for System to be Unstable is that
• All Signs of Coefficients of Denominator of Closed-Loop Transfer
Function are not the Same
• If Powers of s are missing, System is
• either Unstable
• or Marginally Stable
Automated Controls
MEN 435 Sheet 25
Prof. Dr.-Ing. Saleh M. Chehade
 Introduction
• Unfortunately
• if all Coefficients of Denominator are
• Positive and
• not Missing
• we do not have definitive Information about
System’s Pole Locations
• If the described Method is not sufficient, then
• Computer can be used to determine Stability
• by calculating Root Locations of Denominator of Closed-Loop
Transfer Function

Automated Controls
MEN 435 Sheet 26
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
• In the next Slides, we learn
• Method that yields Stability Information without Need to solve for
Closed-Loop System Poles
• Using this Method
• we can tell how many Closed-Loop System Poles are in
• Left Half-Plane
• Right Half-Plane and
• on j -axis
• We can find Number of Poles in each Section of s-Plane
• but we cannot find
their Coordinates
Automated Controls
MEN 435 Sheet 27
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion

Automated Controls
MEN 435 Sheet 28
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
• The Method is called Routh-Hurwitz Criterion for Stability
• Method requires Two Steps
1. generate Data Table called Routh Table and
2. interpret Routh Table to tell how many Closed-Loop System
Poles are
a) in Left Half-Plane
b) in Right Half-Plane, and
c) on j -axis
• You might wonder
• why we study Routh-Hurwitz Criterion
• when Computers can tell the Exact Location of System Poles
Automated Controls
MEN 435 Sheet 29
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
• The Power of Method
• lies in Design
• rather than Analysis
• For Example
• if you have Unknown Parameter in Denominator of Transfer
Function
• it is difficult to determine via Calculator/Computer
• Range of this Parameter to yield Stability
• you would probably rely on Trial and Error
• to answer Stability Question

Automated Controls
MEN 435 Sheet 30
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion
• Generating Basic Routh Table

Automated Controls
MEN 435 Sheet 31
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• Look at Equivalent Closed-Loop Transfer Function shown in Figure

• Since we are interested in System Poles

• focus Attention on Denominator
• First
• create Routh Table as per Next Slides

Automated Controls
MEN 435 Sheet 32
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• begin by labelling Rows with Powers of s from Highest Power of
Denominator of Closed-Loop Transfer Function to s0
• Next
• start with Coefficient of Highest Power of s in Denominator and
• list, horizontally in First Row, every other Coefficient

Automated Controls
MEN 435 Sheet 33
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• In Second Row,
• list horizontally, starting with Next Highest Power of s
• every Coefficient that was skipped in First Row

Automated Controls
MEN 435 Sheet 34
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• Remaining Entries are filled in as follows
• Each Entry is Negative Determinant of Entries in Previous two Rows
divided by Entry in First Column directly above Calculated Row

Automated Controls
MEN 435 Sheet 35
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• Remaining Entries are filled in as follows
• Each Entry is Negative Determinant of Entries in Previous two Rows
divided by Entry in First Column directly above Calculated Row

Automated Controls
MEN 435 Sheet 36
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• The Left-Hand Column of Determinant is always First Column of
Previous Two Rows
• The Right-Hand Column is Elements of Column above and to Right

Automated Controls
MEN 435 Sheet 37
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Generating Basic Routh Table
• Table is Complete when all of Rows are completed down to s0
• The Table below is completed Routh Table

Automated Controls
MEN 435 Sheet 38
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Example – Generating Basic Routh Table
• PROBLEM: Make Routh table for System Shown in Figure

Automated Controls
MEN 435 Sheet 39
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Example – Generating Basic Routh Table

Automated Controls
MEN 435 Sheet 40
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion
• Generating Basic Routh Table
• Interpreting Basic Routh Table

Automated Controls
MEN 435 Sheet 41
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Interpreting the Basic Routh Table
• Now that we know how to generate Routh Table

• let us see how to interpret it

• The Basic Routh Table applies to Systems with Poles in

• Left and Right Half-Planes

• Systems with Imaginary Poles and Kind of Routh Table that results

• will be discussed Later

Automated Controls
MEN 435 Sheet 42
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Interpreting the Basic Routh Table
• Simply stated, Routh-Hurwitz Criterion declares that

• Number of Roots of Polynomial that are in Right Half-Plane

is equal to

• Number of Sign Changes in First Column

• If Closed-Loop Transfer Function has all Poles in Left Half of s-Plane

• System is Stable

• Thus, System is Stable if

• there are no Sign Changes in First Column of Routh Table

Automated Controls
MEN 435 Sheet 43
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Interpreting the Basic Routh Table

• For Example, Table from Example

• has Two Sign Changes in 1st First Column
• 1st Sign Change occurs from 1 in s2 Row to -72 in s1 Row
• 2nd occurs from -72 in s1 Row to 103 in the s0 Row
Automated Controls
MEN 435 Sheet 44
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Interpreting the Basic Routh Table
• Thus,
• Below System is Unstable since
• 2-Poles exist in Right Half-Plane

Automated Controls
MEN 435 Sheet 45
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion
Skill Assessment
PROBLEM: Make Routh Table and tell how many Roots of the Following
Polynomial are in Right Half-Plane and in Left Half-Plane

Automated Controls
MEN 435 Sheet 46
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion
• Generating Basic Routh Table
• Interpreting Basic Routh Table
3. Routh-Hurwitz Criterion : Special Cases

Automated Controls
MEN 435 Sheet 47
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
…
• Two Special Cases can occur
1. Routh Table sometimes will have
• Zero only in First Column of a Raw
2. Routh Table sometimes will have
• Entire Row that consists of Zeros

Automated Controls
MEN 435 Sheet 48
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion
• Generating Basic Routh Table
• Interpreting Basic Routh Table
3. Routh-Hurwitz Criterion : Special Cases
• Zero Only in First Column

Automated Controls
MEN 435 Sheet 49
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Zero Only in First Column
• If 1st Element of Row is Zero
• division by Zero would be required to form Next Row
• To avoid this Phenomenon
• Epsilon is assigned to replace Zero in First Column
• Value is then allowed to approach Zero from
• either Positive Side
• or Negative Side
after which
• Signs of Entries in First Column can be determined

Automated Controls
MEN 435 Sheet 50
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Example : Stability via Epsilon Method
• PROBLEM: Determine Stability of Closed-Loop Transfer Function

Automated Controls
MEN 435 Sheet 51
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Example : Stability via Epsilon Method
• PROBLEM: Determine Stability of Closed-Loop Transfer Function

Automated Controls
MEN 435 Sheet 52
Prof. Dr.-Ing. Saleh M. Chehade
 Topics

• This Presentation
1. Introduction
2. Routh-Hurwitz Criterion
• Generating Basic Routh Table
• Interpreting Basic Routh Table
3. Routh-Hurwitz Criterion : Special Cases
• Zero Only in First Column
• Entire Row in Zero

Automated Controls
MEN 435 Sheet 53
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• Now look at 2nd Special Case
• Sometimes while making Routh Table
• we find that Entire Row consists of Zeros because there is an
Even Polynomial that is a Factor of Original Polynomial
• This Case must be handled differently from
• Case of Zero in only 1st Column of Row
• Next, look at an Example that demonstrates how to
• construct and
• interpret
Routh Table when an Entire Row of Zeros is present

Automated Controls
MEN 435 Sheet 54
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Example – Stability via Routh Table with Row of Zeros
PROBLEM: Determine Number of Right-Half-Plane Poles in Closed-Loop Transfer
Function

• Build a Polynomial from the Row above

• Take Derivative

• Replace Zero Raw with

Coefficients of Derivative
• Table shows no change in Signs accordingly no Roots in Right Hand Plane

Automated Controls
MEN 435 Sheet 55
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• Let us look further into the Case that yields
• Entire Row of Zeros
• Entire Row of Zeros will appear in Routh Table when
• Purely Even or
• Purely Odd Polynomial
is Factor of Original Polynomial

Automated Controls
MEN 435 Sheet 56
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• For Example, s4 +5s2+7 is Even Polynomial
• it has only Even Powers of s
• Even Polynomials only have
• Roots that are Symmetrical about Origin
• This Symmetry can occur under 3 Conditions of Root Position
1. Roots are Symmetrical and Real
2. Roots are Symmetrical and Imaginary or
3. Roots are Quadrantal

Automated Controls
MEN 435 Sheet 57
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• Figure shows Examples of mentioned Cases
• Each Case or Combination of these Cases will generate an
• Even Polynomial
• It is this Even Polynomial that causes
Row of Zeros to appear
• Thus,
• Row of Zeros tells the Existence of
Even Polynomial whose Roots are
Symmetric about Origin
• Some of these Roots could be on j -Axis

Automated Controls
MEN 435 Sheet 58
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• On Other Hand
• since j Roots are Symmetric about
Origin
• if we do not have Row of Zeros
• we cannot possibly have
• j Roots

Automated Controls
MEN 435 Sheet 59
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
• Another Characteristic of Routh Table for Case in Question is that
• Row previous to Row of Zeros contains
• Even Polynomial that is a Factor of Original Polynomial
• Finally everything
• from Row containing Even Polynomial down
• to End of Routh Table is
• Test of only Even Polynomial

Automated Controls
MEN 435 Sheet 60
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
Example – Pole Distribution via Routh Table with Row of Zeros

Automated Controls
MEN 435 Sheet 61
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
Example – Pole Distribution via Routh Table with Row of Zeros

Automated Controls
MEN 435 Sheet 62
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
Example – Pole Distribution via Routh Table with Row of Zeros

Analysing Table from s4 and below to s0

• Polynomial above appears to be Even, meaning
• Roots are Symmetric
• From s4, Table shows no change in Signs in the First Column,
meaning
• no Roots in the RHP
• Since no Roots in RHP, the no Roots in LHP to sustain Symmetry,
meaning
• The 4 Roots could only be on the Imaginary Axis
• These Conclusions are summarized in the next Table

Automated Controls
MEN 435 Sheet 63
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
Example – Pole Distribution via Routh Table with Row of Zeros
• Analyzing Table from s8 down to s4
• There 2 Changes in Sign
• From s7 to s6 and
• From s6 to s5
• Meaning two Roots in RHP
• These Conclusions are summarized in Table

Automated Controls
MEN 435 Sheet 64
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Entire Row is Zero
Example – Pole Distribution via Routh Table with Row of Zeros

Automated Controls
MEN 435 Sheet 65
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Special Cases
Skill Assessment Exercise
PROBLEM: Use Routh-Hurwitz Criterion to find how many Poles of Following
Closed-Loop System, T(s), are in rhp, in lhp, and on j -axis:

Automated Controls
MEN 435 Sheet 66
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Exercise – Standard Routh-Hurwitz
PROBLEM: Find Number of Poles in the Left Half-Plane, Right Half-Plane, and on
j -axis for System

Automated Controls
MEN 435 Sheet 67
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Exercise – Routh-Hurwitz with Zero in First Column
• PROBLEM: Find Number of Poles in
• Left Half-Plane
• Right Half-Plane and
• on j
for System in

Automated Controls
MEN 435 Sheet 68
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example – Routh-Hurwitz with Row of Zeros
• PROBLEM: Find Number of Poles in
• Left Half-Plane,
• Right Half-Plane, and on
• j -axis
for System in Figure
• Draw Conclusions about Stability of Closed-Loop System

Automated Controls
MEN 435 Sheet 69
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
• Routh-Hurwitz Criterion gives vivid Proof that
• Changes in Gain of Feedback Control System result in
• Differences in Transient Response because of
• Changes in Closed-Loop Pole Locations

• For Control Systems, such as those

shown in Figure
• Gain Variations can move
Poles
• from Stable Regions of s-
Plane onto j -Axis and
• then into Right Half-Plane
Automated Controls
MEN 435 Sheet 70
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz
• Find Range of Gain, K, for System of Figure that will cause System to be
• Stable
• Unstable and
• Marginally Stable
• Assume K > 0

Automated Controls
MEN 435 Sheet 71
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz

Automated Controls
MEN 435 Sheet 72
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz

• Since K is assumed Positive

• we see that all elements in First Column are always positive except Row s1
• This entry can be
• Positive, Zero, or Negative
depending upon Value of K.
• If K < 1386
• all terms in the 1st Column will be Positive, and
• since there are no Sign Changes,
• System will have 3 Poles in LHP and be Stable
Automated Controls
MEN 435 Sheet 73
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz

• If K > 1386
• s1 Term is Negative, and
• there are 2 Sign Changes, indicating System has
• 2 Poles in RHP
• 1 Poles in LHP
and System is Unstable

Automated Controls
MEN 435 Sheet 74
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz

• If K = 1386
• Entire Row of Zeros
• which could signify j Poles
• Returning up One Row to s2 and building the Polynomial

Automated Controls
MEN 435 Sheet 75
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example - Stability Design via Routh-Hurwitz

• Conclusion
• No Sign Change from s2 and below (no RHP)
• since Polynomial is Even sustaining Symmetry (no LHP either)
• Poles are on j -Axis
• No Sign Change above s2 (no RHP Poles)
• since Polynomial is Even sustaining Symmetry (no LHP either)
• Poles are on j -Axis
• System is Marginally Stable
Automated Controls
MEN 435 Sheet 76
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example – Factoring via Routh-Hurwitz
• Problem : Factor the Polynomial

Automated Controls
MEN 435 Sheet 77
Prof. Dr.-Ing. Saleh M. Chehade
 Routh-Hurwitz Criterion: Additional Examples
Example – Factoring via Routh-Hurwitz
• Problem : Factoring the Polynomial

• This Polynomial is differentiated with respect to s in order to complete the Routh

table.
• However, since this Polynomial is a Factor of Original Polynomial
• dividing Original Polynomial by this One, yields the other Factor
(s2 + 3s + 20)
• Hence

Automated Controls
MEN 435 Sheet 78
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
• Up to this Point we have examined Stability from s-Plane Viewpoint
• Now we look at Stability from Perspective of State Space
• In Previous Presentation, we mentioned that
• Values of System’s Poles are equal to
• Eigenvalues of System Matrix A
• We stated that Eigenvalues of Matrix A were Solutions of Equation
det(sI - A) = 0
which also yielded Poles of Transfer Function
• Let us now formally show that
• Eigenvalues and
• System Poles
have same Values

Automated Controls
MEN 435 Sheet 79
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
…..
• Eigenvalues of Matrix A, are values of that permit
• Nontrivial Solution (other than 0) for x in Equation
Ax = x 6.26

• In order to solve for Values of that do indeed permit Solution for x

• We rearrange Eq. (6.26) as follows
x – Ax = 0 6.27

( I – A)x = 0 6.28

x = ( I – A)-10 6.29

6.30

Automated Controls
MEN 435 Sheet 80
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space

6.30

• All Solutions will be Null Vector except for

• Occurrence of Zero in Denominator
• Since this is the Only Condition where Elements of x will be
• 0/0, or
• Indeterminate
it is the only Case where Nonzero Solution is possible
• Values of x are calculated by forcing Denominator to Zero
det( I – A) = 0 6.31

Automated Controls
MEN 435 Sheet 81
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
Ax = x 6.26

det( I – A) = 0 6.31
• This Equation determines
• Values of x for which Nonzero Solution for x in Eq. (6.26) exists.
• In previous Presentation, we defined
• x as Eigenvectors and
• Values of x as Eigenvalues
of Matrix A

Automated Controls
MEN 435 Sheet 82
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
3.73

• Let us now relate Eigenvalues of System Matrix, A, to System’s Poles

• In previous Presentation,
• we derived Equation of System Transfer Function, Eq. (3.73), from
State Equations
• The System Transfer Function has
det( I – A)
in Denominator because of Presence of ( I – A)-1
• Thus
det( I – A) = 0 6.32

is Characteristic Equation for System from which System Poles can be found
Automated Controls
MEN 435 Sheet 83
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
det( I – A) = 0 6.31 det( I – A) = 0 6.32
• Since Eqs. (6.31) and (6.32) are identical apart from Change in Variable
Name
• we conclude that Eigenvalues of Matrix A are identical to System’s
Poles
• before cancellation of
• Common Poles and Zeroes in Transfer Function
• Thus, we can determine Stability of System represented in State Space
• finding Eigenvalues of System Matrix, A, and
• determining their Locations on s-Plane

Automated Controls
MEN 435 Sheet 84
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
Example – Stability in State Space
PROBLEM: Given the system

• find out how many poles are

• in left half-plane,
• in right half-plane, and
• on j -Axis

Automated Controls
MEN 435 Sheet 85
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
Example – Stability in State Space
PROBLEM: Given the system

• Then finding

• Building the RHT

Automated Controls
MEN 435 Sheet 86
Prof. Dr.-Ing. Saleh M. Chehade
 Stability in State Space
Example – Stability in State Space
PROBLEM: Given the system

• Since there is One Sign Change in First Column, System has

• One RHP Pole and
• Two LHP Poles
• It is therefore Unstable

Automated Controls
MEN 435 Sheet 87
Prof. Dr.-Ing. Saleh M. Chehade