See discussions, stats, and author profiles for this publication at: [Link]

net/publication/283580045

Comparison Study of Different Structures of PID Controllers

Article in Research Journal of Applied Sciences, Engineering and Technology · July 2015
DOI: 10.19026/rjaset.11.2026

CITATIONS READS
0 6,022

4 authors, including:

Abdel Aziz Mahmoud Abouelfotouh Mona Amin

National Research Centre National Research Centre
6 PUBLICATIONS 14 CITATIONS 49 PUBLICATIONS 575 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Mona Amin on 09 November 2015.

The user has requested enhancement of the downloaded file.

Research Journal of Applied Sciences, Engineering and Technology
ISSN:
© Maxwell Scientific Organization, 2015
Submitted: Accepted: Published:

Comparison Study of Different Structures of PID Controllers

Ahmed S. Abd El-Hamid, Ahmed H. Eissa, A.M. Abouel-Fotouh and Mona A. Abdel-Fatah
National Research Center, Engineering Division, Dokki, Cairo, Egypt

Abstract: The paper aim to comparative study of different structures of PID controller and the values of gains of the
PID controller Kp, Ki and Kd are calculated by using Levenberg-Marquardt’s Optimization Algorithm (LMA). The
Proportional Integral Derivative (PID) structure is mainly used to achieve the desired output in case of closed loop
control systems in most of the industry applications. It is difficult to obtain the proper values of the PID controller
parameters Kp, Ki and Kd. The performance of this approach is evaluated by setting the objective function as the
integral of the square error (ISE). The proposed methods are simulated by tuning all the parameters of different
structure of PID controller for PH neutralization process in MATLAB program.

Keywords: Chemical process control, cost functions, Levenberg Marquardt’s optimization algorithm, PH
neutralization process, PID controller

INTRODUCTION implementation of modern heuristic optimization

techniques such as genetic algorithms, simulated
PID is a remarkable control strategy widely used in annealing and population based incremental learning
processes industry tries such as oil and gas, chemical, and particle swarm optimization (Kamran et al., 2008).
petrochemical, pulp and paper, food and beverage. PID Recently, Generic Algorithm (GA) has been
controller has been proven in terms of reliability and extensively studied by many researchers in searching
robustness in controlling process variables ranging from for optimal PID parameters due to its high potential of
temperature, level, pressure, flow, pH, etc. Other escaping being trapped at a local minimum (Kada and
factors that attracted industry to choose PID controller Ghazzawi, 2011).
could be due to low cost, easy to maintain, as well as This study is provide a better understanding of how
simplicity in control structure and easy to understand. PID controller is tuned using Levenberg-Marquardt
However, improper PID parameters tuning could lead Algorithm (LMA). The performance of LMA is
to cyclic and slow recovery, poor robustness and the searching local optimal PID parameters and its
worst case scenario would be the collapse of the system reliability to maintain the optimum value for a
operation. This led researchers to explore the beast nonlinear system. This study also compares the
method in searching optimum PID parameters (Astrom transient performance of the pH neutralization system
and Hagglund, 1988). using different structures of PID controller with LMA
Many strategies have been proposed to determine tuning method.
the optimum setting of PID controllers. Ziegler and
Nichols, Cohen and Coon are amongst the pioneers in Levenberg-Marquardt Algorithm (LMA) for tuning
PID tuning methods. They have proposed experimental PID controller: In mathematics and computing, the
PID tuning methods based on trial and error and the Levenberg-Marquardt Algorithm (LMA) is used to
process reaction curve (Vikram et al., 2014). However, solve non-linear least squares problems. These
the difficulties may arise to tune the PID controller minimization problems arise especially in least squares
when the system is complex such as high order, time curve fitting. The LMA interpolates between the Gauss-
delay, non-minimum phase and non-linear processes Newton algorithm (GNA) and the method of gradient
(Yancho Todorov et al., 2013). For example, Ziegler descent. The LMA is more robust than the GNA, which
and Nichols method may give overshoots, highly means that in many cases it finds a solution even if it
oscillatory and longer settling time for a high order starts very far off the final minimum. The LMA is a
system and Cohen-Coon method is only valid for very popular curve-fitting algorithm used in many
systems having S-shaped step response (Ayman, 2011; software applications for solving generic curve-fitting
Fung and Yang, 1998). To overcome these difficulties, problems. However, as for many fitting algorithms, the
various methods have been used to obtain optimum PID LMA finds only a local minimum, which is not
parameters ranging from conventional methods such as necessarily the global minimum (Astrom and
refined Ziegler-Nichols and pole placement and the Hagglund, 1988).

Corresponding Author: Ahmed S. Abd El-Hamid, National Research Center, Engineering Division, Dokki, Cairo, Egypt
1
 Res. J. Appl. Sci. Eng. Technol.,

The primary application of the Levenberg-

Marquardt algorithm is in the least squares curve fitting
problem: given a set of empirical datum pairs of
independent and dependent variables, , optimize
the parameters of the model curve , so that the
sum of the squares of the deviations becomes minimal:

∑ ,

The Levenberg-Marquardt (LM) algorithm is an

iterative technique that locates the minimum of a
multivariate function that is expressed as the sum of
squares of non-linear real-valued functions. In the least- Fig. 1: pH neutralization process
squares problem a function f(x) is minimized that is a
sum of squares:

∑ In all of the above ‘ ’ is the error at time‘ ’.

The error is taken as the input to the PID controller
Problems of this type occur in a large number of which can be expressed as:
practical applications, especially when fitting model
functions to data, i.e., nonlinear parameter estimation.
They are also prevalent in control where you want the
output, , to follow some continuous model
trajectory, for vector and scalar . This problem In this study, the optimum values of the controller
can be expressed as: parameters, which minimize the cost function, are
accurately computed using LMA method. In the pH
, neutralization model, considering PID controller which
is using standard ISE cost function for developing
controlling parameters. Also, different types of
where, , and are scalar functions.
structure of PID controller are tested for this study.
Proper parameter setting is of great importance for a
Performance Index of PID controller: Systems
system to be stable. Having secured a stable system,
performance quantification is achieved through a
cost function must be to adjust the controller parameters
performance index. The performance selected depends
until an optimum response is achieved. Performance
upon the process under consideration and is chosen
indices or cost function have proved to be the most
such that emphasis is placed on specific aspects of the
meaningful measures of dynamic performance. The
systems performance. Alternatively a performance
control parameters to be tuned through the optimization
index is a quantitative measure of systems and is
algorithm are feedback gains of PID of output of the
chosen so that a set of parameters in the systems can be
manipulated variable in the pH neutralization system. In
adjusted to meet the required specification optimally.
this study, the integral of the Square Error (ISE) is
Minimum or maximum value of this index then
chosen as the objective function. The goal of LMA is to
corresponds to the optimum set of parameter value.
seek for minimum fitness value.
Different performance indices used in practice are:
CASE STUDY: PH NEUTRALIZATION
 Integral of the Absolute Error (IAE) where:
PROCESS
| |
Consider a pH neutralization process as shown in
Fig. 1. The flow rates of acid, buffer, base and effluent
 Integral of the Square Error (ISE) where: streams are denoted by , , , respectively.
Output of the process is the pH value of the effluent
stream and the flow rate of base stream, is the
manipulated variable. A dynamic model is derived
 Integral of the Time-Weighted Absolute Error using the conservation laws and reactions equilibrium.
(ITAE) where: The modeling assumptions include perfect mixing,
constant volume of the neutralization tank (V) and
| | complete solubility of the ions involved. The chemical
reactions in the system are as follows:
 Integral of the Time-Weighted Square Error or
Mean Square Error (MSE) where: ↔
2
 Res. J. Appl. Sci. Eng. Technol.,

↔
→
→
→
Fig. 2: Control system
The equilibrium constants for these reactions are:
∝ : Concentration of the ith species in the titrating
stream, mol/L
: Reaction invariant of ith species, mol/L
: Process variable, pH
: Flow rate of the titrating stream, mL/min or
mL/s
The chemical equilibrium equations are modeled Different PID structures: The block diagram of a
using the reaction invariant concept (Gustafsson and simplified control system is shown in Fig. 2. Not all
Waller, 1983). For this system, concentrations of manufactures produce PID’s that conform to the ideal
reaction invariants are defined as: textbook structure. So before commencing tuning it is
important to know the configuration of the PID
algorithm. The majority of text-book tuning rules are
only valid for the ideal architecture. If the algorithm is
different, then the controller parameters suggested by a
particular tuning methodology will have to be altered.
The control signal provided by standard PID
Denoting, the ions neutrality balance in controller is dependent upon three terms and is given
the tank results the following static equation: by:
, 10
10 0 (1)

2 10
1 10 10 where is the error, the controller output and
log , and are the proportional, Integral and
log derivative gains.
The dynamic equations are given by:  Ideal PID structure: The transfer function of the
ideal PID controller is expressed as:

∝ (2) 1

One disadvantage of this ideal configuration is that

a sudden change in set point and hence, the error will
∝ (3) cause the derivative term to become very large and thus
provide a derivative kick to the final control element,
this is undesirable. The derivative mode acts on the
measurement and not the error is an alternative
∝ (4)
implementation. After a change in set point the output
will move slowly avoiding derivative kick after set
where, point changes. This is therefore a standard feature of
: Volume of the mixing tank, mL most commercial controllers. The simulink model of
: Dissociation constant of water, 10-14 ideal PID construe is shown in Fig. 3.
: ith dissociation constant of acid
: Concentration of the ith species in the process  Series (interacting) PID: The transfer function of
stream, mol/L the series PID controller is expressed as:
: Concentration of the ith species in the acid
stream, mol/L
: Concentration of the ith species in the buffer 1 1
stream, mol/L
: Flow rate of acid, buffer and base stream in As with the ideal implementation the series mode
simulation, mL/s can include either derivative on the error or derivative
3
 Res. J. Appl. Sci. Eng. Technol.,

Fig. 3: Simulink model of ideal PID configuration

Fig. 4: Simulink model of series PID configuration

Fig. 5: Simulink model of parallel PID configuration

on the measurement. The simulink model of series PID  Parallel PID: The transfer function of the parallel
is shown in Fig. 4. PID controller is expressed as:
4
 Res. J. Appl. Sci. Eng. Technol.,

SIMULATION RESULTS AND DISCUSSION

The pH neutralization closed loop control system

The proportional gain only acts on the error, was solved using Levenberg-Marquardt’s optimization
whereas with the ideal algorithm it acts on the integral approach with sampling time of 0.001 s. The simulation
and derivative modes as well. The simulink model of method combines SIMULINK module and M-file
parallel PID construe is shown in Fig. 5. where, the main program is realized in SIMULINK and

Fig. 6: Step response and PID parameters values of ideal PID configuration

Fig. 7: Step response and PID parameters values of ideal PID configuration
5
 Res. J. Appl. Sci. Eng. Technol.,

the optimized PID controller is predicted using M-file. PID controller. The comparison among different
The closed loop response of the pH neutralization structure of PID controller in terms of various
transfer function subjected to a step input with different performance specification such as rise time, settling
methods has been shown in Fig. 6 to 11 using different time, overshoot and steady state error using Levenberg-

Fig. 8: Step response and parameters values of series PID configuration

Fig. 9: Step response and parameters values of series PID configuration

6
 Res. J. Appl. Sci. Eng. Technol.,

Fig. 10: Step response parameters values of parallel PID configuration

Fig. 11: Step response parameters values of parallel PID configuration

Table 1: Comparison among different PID controllers

LMA Tuning Method
-------------------------------------------------------------------------------------------------------------------
Performance specification parameters Ideal PID controller Series PID controller Parallel PID Controller
Rise time 40 min 45 min 30 min
Settling time 50 min 70 min 65 min
Overshoot 10 % Zero Zero
Steady State error Zero 0.5 % Zero

7
 Res. J. Appl. Sci. Eng. Technol.,

Marquardt’s optimization approach has been shown in Fung, R.F. and R.T. Yang, 1998. Application of
Table 1. The various performance specifications have variable structure controller in position control of a
been ideal PID controller, the best performance in terms nonlinear electrohydraulic servo system. Comput.
of settling time and overshoot has been given by Struct., 66(4): 365-372.
Parallel PID controller. Gustafsson, T.K. and K.V. Waller, 1983. Dynamic
modelling and reaction invariant control oh PH.
CONCLUSION Chem. Eng. Sci., 38: 389-398.
Kada, B. and Y. Ghazzawi, 2011. Robust PID
A comparison between the different structures of controller design for an UAV flight control system.
PID controller is presented. Optimization of the PID Proceedings of the World Congress on Engineering
controller gains for a pH neutralization systems using and Computer Science (WCECS 2011), Vol II, San
LMA technique has been proposed. The LMA Francisco, USA.
technique is utilized to evaluate the PID controller gains Kamran, T., O. Ugur, [Link] and M.J. Elbrous, 2008.
which improve the dynamic performance of the pH PID parameter optimization of an UAV
neutralization system to an operating condition with longitudinal flight control system. Int. J. Mech.
perturbations.
Aerosp. Ind. Mech. Manuf. Eng., 2(9): 35-40.
Vikram, C., K.S. Sunil and D. Lillie, 2014.
REFERENCES
Comparative analysis of tuning a PID controller
Astrom, K.J. and T. Hagglund, 1988. Automatic Tuning using intelligent methods. Acta Polytech. Hung.,
of PID Controllers. Instrument Society of America, 11(8): 235-249.
Research Triangle Park, NC.
Ayman, A.A., 2011. PID parameters optimization using Note: Provide us missing information of following
genetic algorithm technique for electrohydraulic reference which is mentioned in text but not
servo control system. Intell. Control Automat., 2: available in reference list.
69-76. Yancho Todorov et al. (2013).

View publication stats