International Journal of Computer Applications (0975 – 8887)

Volume 103 – No.14, October 2014

Application of various PID Controller Tuning Techniques

for a Temperature System
V. Selvi Sowmya S. Priya dharsini R.Priya Dharshini P. Aravind
UG student, UG student, UG student, Assistant Professor,
Department of ICE Department of ICE Department of ICE Department of ICE
Saranathan College Saranathan College Saranathan College Saranathan College
of Engineering, of Engineering, of Engineering, of Engineering,

ABSTRACT The PID controller is very popularly used in the feedback

Control of temperature level in the process tank is vital in network. It has good clarity and its easy to implement. A PID
process industries. A proportional-integral-derivative controller helps to bring down the difference between the
controller (PID controller) is a control loop feedback process variable and the set point by outputting the response
mechanism (controller) widely used in industrial control with the desired value. PID controller has zero steady state
systems.PID control is a control strategy that has been error, fast response, short rise time, no oscillations and higher
successfully used over many years. This algorithm involves stability. PID controllers are preferred over PI controllers
three separate constant parameters. And it is otherwise called because they reduce the overshoot and it also holds added
as three term controller. Where, P depends on the present advantage of employability for higher order systems. A PID
error, I on the accumulation of past errors, and D is a controller propels a control signal that has a component
prediction of future errors, based on current rate of change. It proportional to the error of a system, accumulation of the
is highly used in chemical industries because of their success error over time and the rate of change of the error with respect
rate in practical applications, robustness and simplicity The to time.
PID temperature control algorithm is used for the control of Mostly every system will have multiple and contradictory
almost all loops in the process industries, and is also the basis objectives to be achieved. For designing a controller by
for many advanced control algorithms and strategies. A PID satisfying all the requirements we need algorithms so as to
temperature controller works by measuring the current tackle the problems that may arise. The conventional tuning
temperature to generate the error signal and altering the output methods which works based on fixed parameters will result in
to bring it closer to the target. The main alternative for lesser performance when system necessitates controller. We
temperature PID controller is ON/OFF controller which works will discuss about the four tuning techniques-Ziegler-Nichols
like a thermostat. To determine the appropriate controller (ZN), Ziegler Chien Hrones Reswich (C-H-R), Tyreus-
parameters settings, a conventional PID control tuning lubeyn(T-L) for designing the PID controllers. From the
techniques and model based PID were characterized and their above specified tuning methods, the proportional band,
results to be obtained are simulated. With the yield of time integral time and derivative time can be calculated. By using
domain and error criterion analysis, the most suitable those values one can determine the Proportional constant
controller for temperature controlled process is obtained. The (Kc), Integral constant (Ki) and Derivative constant (Kd).It
best among the tuned controlled technique is pointed out as a also includes the PID values [Kc,Ki,Kd] of the four tuning
good result of minimum rise time, setting time and overshoot. methods and the Tuning method of Minimum Error Integral
Criteria for determining the error values of ITAE, ISE, IAE,
General Terms MSE. Temperature control is a process in which change of
Process Control, Control System. temperature of a space (and objects collectively there within)
is measured or otherwise detected, and the passage of heat
Keywords energy into or out of the space is adjusted to achieve a desired
PID controller, Temperature Control, CHR method. average temperature.
1. INTRODUCTION 2. SYSTEM IDENTIFICATION
In 1930s Proportional - Integral – Derivative (PID) controller
came to exist. Because of it generic and efficient output for 2.1. Mathematical Modeling
real world control problems; it is being used even today. To analyze the behavior of a process, a mathematical
Hence it has been widely accepted in major process industries. representation of the physical and chemical phenomenon
It makes things easier for the operator who has to look into taking place in it is essential and this representation
control of the system. Different tuning methods are compared constitutes the mathematical model.
in this study. To achieve optimum performance following are
essential, 2.2. Experimental Setup
The experimental set up shown in Figure 1 consists of water
1. Dynamic model of the system, supply tank, pump, Plunger pump, temperature transmitter,
pressure transmitter, Process tanks with differential pressure
2. Desired performance of closed loop system on the basis of
transmitters, rotameter, pneumatic control valve, I/P
known physical parameters,
converter, Pressure gauges(16). The pneumatic control valve
3. Adopting controller strategies for desired performance is air to close, adjusts the flow of the water pumped to the
cylindrical tank from the water reservoir. The temperature of
4. Implementation of the controller using suitable platform, the water in the tank is measured by means of the temperature
5. Validation of controller performance (1) transmitter and is transmitted in the form of (4-20) mA to the

32
 International Journal of Computer Applications (0975 – 8887)
Volume 103 – No.14, October 2014

interfacing ADAM’s module to the Personal Computer (PC). The proportional, integral and derivative values of tyreus-
In this process, either RTD (Resistance Temperature Detector) lubeyn method are Kp=0.124, Ki=0.00773, Kd=0.143.The
or Thermocouple can be used to measure the temperature response for this method has a rise time 235 second and it
level in the process. A step test is performed on temperature settles at 235 second (11).
process station and process is identified as first order plus
dead time model defined by terms process gain(K)-5, time 3.3. Chien-Hrones-Reswick Method:
constant(𝜏)-2, dead time (Td)-4. The Chien-Hrones-Reswick auto tuning method focuses on
setpoint response and disturbance response. This method
provides formulas for 0% and 20% overshoot (18).
Table 1: Controller Tuning Parameters
Controller Kc Ti Td
P 0.3TP/ — —
PI 0.6TP/ 4 —
PID 0.95TP/ 2.4 0.42
where TP is the time constant.
The proportional, integral and derivative values Chien-
Hrones-Reswick of are Kp=0.12, Ki=0.015, Kd=0.20.The
Fig. 1Experimental Setup of Temperature Process
response for this method has a rise time 98 second and it
3. PID TUNING METHODS settles at 98 second.
The various tuning methods used here are 4. ERROR CRITERIA
1. Zieger-Nichols (Z-N) method 4.1. Integral of the Absolute Value of the
2. Tyreus-Lubeyn method(T-L) method Error (IAE):
Integral absolute error is a process which fits a sample to a
3. Chien-hrones-Reswick (C-H-R) method certain outcome to show the closeness of the fitness measure
of the results to the targeted value (18). Its response is slower
3.1. Zieger-Nichols Method than ISE but with less sustained oscillation.
The Ziegler–Nichols tuning method is a heuristic method of
tuning a PID controller. It was developed by John G. Ziegler T
and Nathaniel B. Nichols. It is performed by setting the I
(integral) and D (derivative) gains to zero.(4) The "P" IIAE = e(t) dt
(proportional) gain, Kp is then increased (from zero) until it 0
reaches the ultimate gain, ku at which the output of the control
loop oscillates with constant amplitude. Ku and the oscillation 4.2. Integral of the Square Value of the
period Tu are used to set the P, I, and D gains depending on Error (ISE):
the type of controller used (6). Integral square error is error from an output, squared and
The Ziegler-Nichols method attempts to produce good values added integrated over time in continuous systems is used to
for the three PID gain parameters: measure system performance in applications of optimal
control and estimation(4). It eliminates the large errors
1. Kp - the controller path gain quickly. It produce fast responses, but with considerable, low
amplitude, oscillation.
2. Ti - the controller's integrator time constant
T
3. Td - the controller's derivative time constant
IISE = e2 (t) dt
The proportional, integral and derivative values of Ziegler-
Nichols method are Kp=0.238,Ki=0.065, Kd=0.2168. The 0
response for this method has a rise time 5.66 second and it
attains the overshoot of 1.17 % at 7.58 second and it settles at 4.3. Integral of the Time Weighted Absolute
22 second. Value of the Error (ITAE):
3.2. Tyreus-Lubeyn Method In ITAE absolute error multiplied by time is integrated over
This method (18) is also called as online tuning method. They time. It is to weight errors which exist after a long time much
had developed tuning method in closed-loop mode. This more heavily than those at the start of the response. Systems
closed-loop tuning method overcomes the shortcoming of the tuned by ITAE will settle much more quickly when compared
well-known Ziegler-Nichols continuous cycling method and to other two tuning methods (15). It produces system with
gives consistently better performance and robustness for sluggish initial response which avoids sustained oscillation.
T
broad class of processes. The PI/PID controller design method
has been discussed extensively in the literature and it shows IITAE = t e(t) dt
that most of the tuning method is based on the two steps
procedure which is very easy to calculate the parameters. 0

33
 International Journal of Computer Applications (0975 – 8887)
Volume 103 – No.14, October 2014

4.4. Mean Square Error (MSE): 6. CONCLUSION

Mean Square Error is simple to use, natural to define energy The model is characterized and control system for temperature
of a signal, which possesses symmetry and differentiability process is analyzed using MATLAB platform. Proportional
(4). It is widely used in signal processing. band, Derivative time and Integral time are fed to PID
controller. Output is recorded and the results of various tuning
Table 2. PID Tuning Values methods depict that Chien-Hrones-Reswick (CHR) tuned PID
Tuning Methods Kp Ki Kd controller is better than other PID tuning techniques
compared. The simulated response for the model corroborated
Ziegler Nichol’s 0.238 0.065 0.2168 shows the effectiveness of CHR PID controller in terms of
time domain specification. The performance index of CHR
Tyreus Luyben 0.124 0.00773 0.143
method PID controller is also better than other PID controllers
CHR 0.12 0.015 0.20 compared from the real time responses; the CHR method PID
controller is suitable for this temperature process.
5. RESULT AND COMPARSION 7. REFERENCES
1.4 [1] Mohammad Shahrokhi and AlirezaZomorrodi, (2012)
Comparison of PID Controller Tuning Methods
1.2
[2] Astrom K, J, T. Hagllund, (1994), PID controller Theory,
1 zn Design and Tuning, 2ndedition, Instrument Society of
0.8 America
TL
Temperature

0.6 [3] Chen C.L., (1989), a Simple Method for Online

CHR Identification and Controller Tuning
0.4
[4] Coughanowr D.R. (1991), Process System Analysis and
0.2 Control,2ndedition McGrawHill
0 [5] Erickson K.T., J.L. Hedrick, (1999), Plant wide Process
Control John Wiley & Sons
-0.2 0 100
Time in sec 200 300
[6] Hang C.C., J.K. Astrom, W.K. Ho, (1991) Refine-ments
-0.4 of Ziegler Nichols Tuning formulaIEE Proceedings
-0.6 [7] Krishnaswamy P.R., B.E Mary Chan., and G.P,
Rangaish; (1987) Closed- Loop Tuning of Control
Figure 2: Comparison of Simulated Response for different Systems, Chem. Eng.Sci.42, 2173
control techniques [8] Luyben W.L, M.L. Luyben; (1997) Essentials of Process
The Comparisons of different PID controllers tuning methods Control, McGraw-Hill
are below, [9] Marlin T.E. ;(2000) Process Control, 2 nd edition, ,Mc
graw -Hill
Table 3. Time Domain Specification:
[10] Morari M., E. Zafirion; (1989) Robust Process Control,
Tuning Peak Rise Peak Settling Prentice-Hall, EnglewoodCliffs, NJ
Methods Time Time Overshoot Time
[11] Ogata K.; (1997) Modern Control Engineering, 3rd
Ziegler 7.58 5.66 1.117 22 edition, Prentice Hall
Nichol’s [12] Seborg, D.E., T.F. Edgar, D.A.Mellichamp; (1989)
Process Dynamics and Control, John Wiley & sons
Tyreus - 235 - 235
Luyben [13] Smith, C.A., A.B. Copripio; (1985) Principlesand
Practice of Automatic Process Control, John Wiley &
Chien - 98 - 98 Sons Stephanopolous, G.; (1984) Chemical Process
Hrones Control, Prentice-Hall
Control
[14] Yuwata, M, D.E. Seborg; (1982) A new Method for
Online Controller Tuning
The Comparison of different PID Controllers ITAE, IAE, ISE,
MSE performance index are given below [15] K.Mohamed Hussain, R.AllwynRajendraZepherin,
M.Shanha Kumar, Comparison of PID controller tuning
Table . 4. Error Response Criteria: methods for FOPTD system.
Tuning ISE IAE ITAE MSE [16] A.R. Laware, V.S. Bandal and D.B. Talange,(2013), Real
Methods Time Temperature Control System Using PID controller
and Supervisory Control and Data Acquisition System
Ziegler 63.8401 1.4727 1.1395 0.0044 (SCADA), International Journal of Application or
Nichol’s E+006 E+005 Innovation in Engineering &Management(IJAIEM),\
Tyreus Luyben 0.29999 1.0866 2.0116 0.0413 [17] S.M. Giriraj Kumar, R. Ravishankar, T.K. Radha
E+005 E+005 Krishnan, V. Dharmalingam and N.
Anantharaman,(2008), Particle Swarm Optimization
Chien Hrones 181.576 1.8759 1.9474 0.0051 Technique Based Design of PI Controller for a Real time
control E+005 E+005 Non- Linear Process, Instrumentation Science and
Technology.

IJCATM : www.ijcaonline.org
34