International Journal of Pure and Applied Mathematics

Volume 118 No. 10 2018, 107-114

ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version)
url: http://www.ijpam.eu
doi: 10.12732/ijpam.v118i10.14
Special Issue ijpam.eu

DESIGN OF ADAPTIVE PID CONTROLLER FOR CONTINUOUS STIRRED

TANK HEATER (CSTH) PROCESS

1
A V Divya Priya, 2G M Tamilselvan, 3T Rajesh
1
PG Scholar, Department of EIE , Bannari Amman Insitute of Technology,
Sathyamangalam-Erode TamilNadu, India.
2
Associate Professor, Department of EIE, Bannari Amman Institute of Technology,
Sathyamangalam-Erode TamilNadu, India.
3
Assitant Professor, Department of EIE, Bannari Amman Institute of Technology,
Sathyamangalam-Erode TamilNadu, India
1
[email protected], [email protected]
3
[email protected]

Abstract: Measurement of temperature is very essential Continuous stirred tank Heater is one of the included
in all process industries. Real-time industrial processes parts of the chemical plant. It has common applications
are focused to variation in parameters determination of in industry and
tuning of the PID (Proportional + Integral + Derivative) a lot of features of additional types of heaters. It is a vital
parameters continues to be significant as these factor of many chemical industries [1]. The study of a
parameters have a huge manipulate on the stability and continuous stirred tank heater (CSTH) is a source for
act of the control system. The need for enhanced the study and control of chemical heaters. Control of its
performance of the process has led to the improvement considerations is a very essential task. It is necessary to
of optimal controllers for the PID's three control keep the volume, temperature and density of the mixed
parameters, Kp, Ki and Kd to obtain the best control of cold water in the desired limits. The distinction may
effect by minimizing the integral square error. So the lead to many terrible things causing injury to the people
control engineers are on seems for automatic tuning and surroundings. A CSTH is a well stirred tank into
procedures. An adaptive controller was determined to which the flows continuously and then the product
develop the performance for the automatic tuning of PID passed out. It can be clear to have a reacting tank and
in CSTH process. For a temperature control process, in that rotates continuously when the process carried out.
the case of adaptive controller PID controllers, the plant CSTH is a container in which the mechanisms are
under concern is assumed to be governed by a set of PID constantly stirred to make certain their suitable mixing
controllers with the aim to obtain good tracking to attain a specific output [2]. The aim is to raise the
behavior over the operating envelope of the plant. The temperature of the inlet stream to a desired value. A heat
PID gains are allowed to differ within a predetermined transfer fluid is spread throughout a jacket to heat the
range. The objective of the work is to design the fluid in the tank. Due to easy implementation and
adaptive PID controller for a temperature control in benefit of PID controller, it is executed in the design of
CSTH process. The real time plant is interfaced with CSTH process to progress the system's transient
LabVIEW with NI data acquisition card. Tuning the response for zero steady state error in addition to
parameters of a controller is necessary for system minimizing the settling time and the peak undershoot
performance. The simulation and real-time execution and overshoot.
were accomplished and its results shown from the
simulation, it is clear that there is significant 2. Literature Survey
enhancement of PID controller in terms of Minimum
Integral Square Error(ISE), Minimum peak overshoot, K.Prabhu et al. proposed the temperature controlfor
Minimum settling time and Minimum rise time. CSTR process using MIT adaptive controller. The
simulation model of CSTR process is modeled in
Keywords: PID, Adaptive control, Integral Square MATLAB and the result provide with minimum
Error(ISE), Process plant, Temperature control overshoot and rise time[4].
Rahul Upadhyay et al. Proposed an adaptive and PID
1. Introduction controller is simulated using MATLAB SIMULINK
program and execution is compared with various
temperature and concentration. They established the
adaptive controller in the occurrence of nonlinearities

107
International Journal of Pure and Applied Mathematics Special Issue

and demonstrates the control of non-linear for

CSTR(Continuous Stirred Tank Reactor)[2]. The term adaptive system means a system with capacity
to do real-time adaptation in response to changes on its
K.Prabhu et al. Proposed a comparison of PID dynamics, usually in the form of parametric variations.
controllers and adaptive controller are very useful An adaptive controller has the responsibility to control a
parameter changes and environmental variations occurs process not only to give suitable response
and it is illustrates clearly in their results. They describes performance[7], and also to permit the entire system to
the behavior of model reference adaptation control using adapt itself to ultimate changes on its dynamics.
MIT rule. By raising the adaptation gain the settling Adaptation mechanisms are finest implemented in terms
time, peak time and rise time was decreased and the of computing algorithms, and therefore adaptive
different time response specification is shown for the controllers are essentially discrete-time controllers[8].
suitable value of adaptation gain in MIT rule and the
plant output tracks the reference model[5]. 3.2 Instrumentation And Utilities
Mohd Fua’ad Rahmat et al. proposed a model os
contin uous stirred tank reactor, which have nonlinear Flow Instruments used to measure the hot & cold water
characteristics. Gain scheduling technique to design the flow rate by orifice plate with differential pressure
adaptive and intelligent control to offer optimum transmitter, processes the output of 4-20mA which is
performance. The result of these controllers provide calibrated to 0-1500 LPH. The valves used in CSTH
accurate performance and robust in functional variations process is pneumatic control valves which operates on
in the plant. The result offered by both controllers are the pressure range of 3-15 Psi Compressed air supply.
more accurate and high performance[3]. The level instruments are used in the CSTH process
S.Boobalan proposed a ZN-PID controller and transmitter in the range of 4-20mA calibrated to 0 to
FUZZY based PID to maintain the set point at desired 60cm of the tank height. The three wires RTD’s are used
value. The performance FUZZY based PID controller is to measure the temperature at different points of process
better when compared to ZN-PID controller. The and produces the output voltage range of 0-5V which is
modeling of process is functioned with operating calibrated to 0 to 250 .
conditions[6].
3.3 Specifications Of Proposed System
3. Proposed System
The System specifications of the tank are mentioned in
3.1 Process Description And Model the Table.I

The plant in the department of Electronics and Table 1. Specification of the Tank
instrumentation engineering at Bannari Amman institute
of technology is the hot and cold water is mixed and Equipments Details
heated using steam through a heating coil in the stirred Stainless steel body,
tank heater. The tank with well stirred and mixed hot height:60cm,
water which implies that the temperature in the tank is CSTH Tank Diameter: 25.78cm.
examined to be the same as the outflow temperature. Shape: Cylindrical
The tank has a cylindrical cross section with a height of Level Transmitter Input:3-15 psi
60 cm and volume 91. The experimental setup for CSTH Output:4-20mA
process is shown in Fig.1 Stirrer Motor RPM:6000,
Frequency:50HZ
Control valve Size 1/4 Pneumatic actuated,
Type: Air to
open,Input:3-15psi
Rotameter Input:3-15psi,
Range:0-460LPH
SCR power pack Output: 4-20mA

3.4 Mass And Energy Balance

Mass balance and energy (heat) balance equations are

formed using algebraic equations which are derived
from experimental data and calibration process are done
for level, Flow transmitters (differential pressure
Figure 1. Experimental setup for CSTH process transmitter), temperature transmitter (3 – wire resistance

108
International Journal of Pure and Applied Mathematics Special Issue

temperature detector(RTD), Valve actuators and Ʈ= = 161.9, Ʈd = 1s

unknown quantities such transfer of heat through the
heating coils[12]. The relationships between these The First Order Process with Time Delay (FOPTD)
parameters are practically non-linear and also the model
constraints are very hard such as tank should not be = 10.57e-1s / 161.9s+1
empty and overflow. … (3)
Mass & Energy balance of the process dynamics are The G(s) in equation (3) is obtained by mathematical
given by the following equations model of CSTH system.

4. Process Related Equations

= fwc + fwh – fo(l) …..(1)
A. Enthalpy
= Wst + hwh + lwhfwh + hwc lwcfwc - houtloutfo(l)
. ….(2) A thermodynamic extent equivalent to the total heat at
Where, ease of a system. It is equal to the internal energy of the
l - Denotes level system boon the product of pressure and volume.
V - Volumes of water H = m* hout (KJ) … (4)
fwh - Hot water flow Where,
fwc - Cold water flow H = Enthalpy of the process (KJ)
fo - outflow m = mass of the system (Kg)
H - Heat rate (total enthalpy) hout = Specific enthalpy (KJ/Kg)
hwh - Specific enthalpy of hot water
hout - Specific enthalpy of outlet water B.Specific Enthalpy
lwc - Cold water density
lwh - Hot water density A thermodynamic equations equivalent to the total
energy in a method temperature per unit of mass in that
The temperature of cold and hot water set to 25 system.
and 55 respectively. For well mixed process
hout = … (4)
3.5 Open Loop Test Where,
H = Overall enthalpy or heat energy of the system
V = Volume of the tank
lout = density of outlet flow from the tank

5. Controller Design

Gain scheduling is shown in the Fig.3 is a method to

control the controllers and it is a common technique of
industrial control and it is used when a plant is focus to
large variation in its operating state, a circumstances that
is characteristic in industry[10]. A gain-scheduled
controller is a controller whose gains are by design
adjusted as a purpose of time, operating condition, or
plant parameters. Gain scheduling is a general strategy
for controlling systems whose dynamics transform with
such variables. The gain scheduling demonstrates about
Figure 2. Open loop Response Curve
the output response towards set point. The output is
implemented to vary the controller gain in a way to
The process reaction curve of this CSTH process is
minimizing the integral square error[11]. The outer loop
attained by performing test of an open loop step test on
is exposed by adaptation method that resembles
the real time process and model parameters for
feedback loop. The standard conventional form of PID
temperature control shown in Fig.2 are recognized
controller is,
from the curve.
For temperature process:
Kp =

= = 10.57

109
International Journal of Pure and Applied Mathematics Special Issue

Figure 3. Adaptive Controller Design

C(S) = Kc (1 + + S) … (9) Figure 5. Block diagram

Kc = Propotional gain
Ʈi = Integration time The error value from the comparator is given to the
Ʈd = Derivative time SCR. The Kc, Ki, Kd values given to the PID controller
For small reference inputs, the response is as good as the is achieved from the Z-N method of controller tuning
one with the gain-scheduled controller. method. From the obtain value the transfer function for
the process is originated from the corresponding voltage
6. Results and Discussion and time seconds values for each °C of temperature of
the RTD. The output from the SCR controller is given to
6.1 Labview Program For Temperature Control the analog output icon (DAQ Assistant 2). When the
DAQ Output (0-5) V, I-P Converter (3-15) PSI,
In this Lab VIEW program, the analog input voltage temperature 45°C set point is reached, 5V is generated in
from the CSTH process station is acquired through the the analog output indicator. For the corresponding
analog input icon from the DAQ (Resistive temperature temperature upto the set point reaches the corresponding
detector (RTD)). The control and simulation (while) intermediate temperature as well as voltages values are
loop used in this program is related to the practical displayed in the Front panel. When the set point is
control loop is Shown in the Block Diagram Fig.5. DAQ reached, the SCR which in turn stops the temperature of
is the device interfaces between the Real time hardware water into the tank.
and the PC. DAQ digitizes incoming analog signals
from the CSTH process station. In the front panel is 6.1.1 Real Time Result For Temperature Control Pid
shown in the Fig.4 set point=45°C is given and SCR to Tuning
stops. The real time process variable is compared with
the given set point 45°C in terms of voltage using a For Temperature control process Set Point=45 and the
comparator. corresponding time the set points reaches is shown in the
Fig.6.

Figure 4. Front Panel Figure 6. Time vs Temperature

110
International Journal of Pure and Applied Mathematics Special Issue

6.2 Adaptive Control Based PID Implementation For

CSTH Process

The control and simulation (while) loop used in this

program is related to the practical control loop is shown
in the block diagram Fig.8. The set point=35cm is given
to the front panel is shown in the Fig.7 and the controller
to stops when the set point reaches. When the DAQ
Output (0-5) V, I-P Converter (3-15) PSI, V-I Converter
(4-20)mA Control Valve 35 set point is reached, 5V is
generated in the analog output indicator. For the
corresponding temperature upto the set point reaches the
corresponding intermediate temperature as well as
Figure 9. Time vs Temperature
voltages values are displayed in the Front panel. When
the set point is reached, the SCR which in turn stops the
6.3 Performance Analysis Of Temperature Control
heating of water into the tank.
The integral error for a step point or disturbance
response:
ISE =
The Simulink model and the corresponding
response curve for Z-N based PID tuning are shown in
the Fig.10.

Figure 7. Front Panel

Figure 10. Simulink Model for Z-N based PID Tuning

The Simulink model and the corresponding

response curve for Z-N based PID tuning are shown in
the Fig .11.

Figure 8. Block Diagram

6.2.1 Real Time Result For Adaptive PID Tuning For

Temperature Control

For Temperature control process Set Point=30 and the

corresponding time the set points reaches is shown in the
Fig.9. In Adaptive control tuning when the set point Figure 11. Simulink Model for cohen-coon
reaches PID controller stops at exact set point. based PID tuning

111
International Journal of Pure and Applied Mathematics Special Issue

Table 2. Comparison of Conventional PID VS Proposed PID

Tuning Gain Integral Derivati Integral Rise Settling Over

Method (KC) Gain ve (Kd) Square Time time shoot
Time (Ki / error (sec) (sec) (%)
Ti) (ISE)
Adaptive PID
controller 8.2 1.4 0.3 22.4 2 20 50%

Ziegler 15.4 3.7 1.3 36.9 5 50 75%

Nicholas

Cohen- 11.2 1.2 0.9 36.9 4 60 88%

coon

From the response curve the time domain

Method” , International Journal of Engineering and
specifications like rise time, % over shoot and settling
InnovativeTechnology
time values were obtained and it is tabulated in table.2.
[5] (IJEIT) Volume 1, Issue 3, March 2012 ISSN:
7. Conclusion 2277-3754
[6] K.prabhu, V.Murali Bhaskaran, “Controller
The simulation and real time experimental results shows design for continuous stirred tank reactor using adaptive
the Adaptive gain scheduling algorithm provides the control” ,Research Journal of Applied sciences9(8):
better output response for the CSTH process when 489-495,2014 ISSN:1815-932X
compare to conventional Ziegler Nicholas and
Cohen-coon tuning methods. The proposed algorithm [7] S.Boobalan, K.Prabhu, V.Murali
progresses the performance of the controller, based on Bhaskaran,”Fuzzy Based Temperature Controller For
set point tracking and reduces the peak overshoot and Continuous Stirred Tank Reactor”International Journal
delay time. From the simulation results it is observed of Advanced Research in Electrical, Electronics and
that the automatic gain scheduling of controller Instrumentation Engineering(An ISO 3297: 2007
parameters are updated within limited period of time. Certified Organization) Vol. 2, Issue 12, December
That the performance of the CSTH process has 2013
improved and the parameters are sustained at their [8] Dipak M.Adhyaru, “State Observer Design For
preferred level of set points with minimum Integral Nonlinear Systems using Neural Network Based HJB
Square Error (ISE), rise time, peak over shoot & settling formulation”, Applied Soft Computing, Vol.12(1), 2012
time.
[9] Adhyaru, D.M.; Kar, I.N.; Gopal, M., “A Fixed
References Final Time Optimal Control Approach for Bounded
Robust Controller Design using HJB Solution”, IET
[1] Middleton R.H., Goodwin, G.C., Hill D.J., and Control Theory and Applications. Vol. 3, No. 9,pp.118
Mayne,D.Q, “Design issues in adaptive control”, IEEE 3-1195, 2009
transaction on automatic control, Vol.33, No.1, [10] Xueli Wu, Xiaojing Wu, Xiaoyuan Luo,
1988,pp.50-58. Quanmin Zhu and Xinping Guan "Neural
[2] Rahul Upadhyay and Rajesh Singla, “Analysis network-based adaptive tracking control for nonlinearly
of CSTR Temperature Control with Adaptive and PID parameterized systems with unknown input
Controller (A Comparative Study)”, IACSIT nonlinearities" eurocomputing, Volume 82, 1 Elsevier,
International Journal of Engineering and Technology, 2012
Vol.2, No.5, October 2010 ISSN: 1793-8236 [11] J. J. E. Slotine, W. Li, “Applied Nonlinear
[3] Mohd Fua’ad Rahmat, Amir Mehdi Yazdani, Control", Prentice-Hall, Englewood Cliffs, 1991.
Mohammad Ahmadi Movahed and Somaiyeh [12] R. Marino and P. Tomei, Nonlinear Adaptive
Mahmoudzadeh,“Temperature control of a continuous DesignGeometric, Adaptive, and Robust, Printice Hall
stirred tank reactor by means two different intelligent International (UK) Limited, London, 1995.
strategies”, “International journal on smart sensing and
intelligent systems” vol. 4, no. 2, june 2011 [13] A.MohammadZadeh, E.Atashpaz Gargari,
C.Lucas, “Vehicle Fuzzy Controller Design Using
[4] K.Prabhu, Dr. V. Murali Bhaskaran,
“Optimization of a Control Loop Using Adaptive

112
International Journal of Pure and Applied Mathematics Special Issue

Imperialist Competitive Algorithm”, IEEE, World

Congress on Computational Intelligence,2008.
[14] Luyben, W. L., “Process Modeling, Simulation
and Control for Chemical Engineers”,McGraw-Hill,
New York, 1989.
[15] Seider, W.D., J.D. Seader and D.R. Lewin
“Process Design Principles”, Wiley (1999).
[16] Adaptive control systems: techniques and
applications By V. V. Chalam:34-98.
[17] Rajesh, M., and J. M. Gnanasekar. "An
optimized congestion control and error management
system for OCCEM." International Journal of
Advanced Research in IT and Engineering 4.4 (2015):
1-10.
[18] S.V.Manikanthan and K.Baskaran “Low Cost
VLSI Design Implementation of Sorting Network for
ACSFD in Wireless Sensor Network”, CiiT
International Journal of Programmable Device Circuits
and Systems,Print: ISSN 0974 – 973X & Online: ISSN
0974 – 9624, Issue : November 2011, PDCS112011008.
[19] T. Padmapriya and V.Saminadan, “Handoff
Decision for Multi-user Multiclass Traffic in
MIMO-LTE-A Networks”, 2nd International
Conference on Intelligent Computing, Communication
& Convergence (ICCC-2016) – Elsevier - PROCEDIA
OF COMPUTER SCIENCE, vol. 92, pp: 410-417,
August 2016.

113
114