International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056

Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072

COMPARATIVE ANALYISIS OF P, PI, PID AND FUZZY LOGIC

CONTROLLER FOR TANK WATER LEVEL CONTROL SYSTEM
Sumit Kumar1, Mrs. Pooja Nagpal2
1JCDM College of Engineering, Haryana, India, Sirsa
2Associate Professor, JCDM college of Engineering, Haryana, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The objectives of this paper are to design and that fuzzy logic control system gives better performance
comparatively analyze the P, PI, PID and Fuzzy Logic than other control systems.
controller for coupled tank liquid level system. Simulink
2. MATHEMATICAL MODELLING OF COUPLED TANK
model for coupled tank system is designed within
SYSTEM
MATLAB/Simulink. Tuning of parameters for PID controller
is done using signal constraint block in MATLAB/simulink.
Let us consider a coupled tank system used in
Simulated results were compared to verify the performance
industrial application as shown in fig.1 the mathematical
of the control system in terms of rise time, steady state
model for this coupled tank system is nonlinear as shown
error, settling time and overshoot. A simulated result of the
by equation 1.1 and 1.2.
controllers in coupled tank system indicates that fuzzy logic
controller gives better results as compared to other
controllers.

Key Words: Coupled tank system, PID controller, Fuzzy

Logic controller, MATLAB/Simulink, Signal constraint.

1. INTRODUCTION

Accuracy in level and flow control of the tanks is

highly important application for the chemical process
industries [1]. Mostly PID controller is used in industries
for control process due to its simple structure and
easiness in implementation [2, 7] Conventional P Fig.1: Schematic diagram of coupled tank system
controller can reduce the adjusting time it can eliminate
the present error only and PID controller use Qin, Q01, Q12 and Q02 are quantity of liquid flow in the tanks,
mathematical model of the system as it is difficult to find quantity of liquid flow out from tank1, quantity of liquid
proper gain so it is not good for highly non linear system flow in between the tanks and quantity of liquid flow out
[3]. In this work coupled tank system is used which is a from tank2.
non linear system and dynamic behavior of tanks affect H1, H2 are the height of the liquid in the tanks, respectively
each other as inflow between the tanks depends on the A1, A2 are the cross sectional area of the tanks, respectively
level of both tanks. Where PI controller is suitable for low α1, α2 and α12 are the cross sectional area of output pipe
order process where accuracy does not the prime factor tank1, tank2 and in between tanks.
but for higher accuracy it is not suitable [4]. It is also Equation for tank 1:-
difficult to get desired result by using conventional PI (A1) = Qin – α1 (√ ) – α12 (√ ) 1.1
controller as it gives poor performance for higher order Equation for tank 2:
process [4]. Fuzzy logic control system does not use (A2) = α12 (√ ) – α2 (√ )
mathematical model of the process so fuzzy logic
controller can be used where PID controller is difficult to 1.2
apply [5].FLC system use human reasoning so it is flexible Using these two equations a simulink model for the
and can be understood easily. It is suitable for the coupled tank has been designed.
requirement of industrial applications it may be a reason A linearization method is applied to get linear equation
behind the wide use of FLC system [6]. So in this paper a and transfer function for the coupled tank system.
comparative analysis of P, PI, PID and FLC has done for Transfer function will be as shown below:
coupled tank system in terms of rise time, setting time,
overshoot and steady state error. The simulated results
are compared and verified in MATLAB/Simulink and find

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1174
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072

Where T1= 2.6085, T2= 2.5839 are the time constants and in surface viewer tool box. When a control rule generates
K1 = 0.081516, K12=0.8934 and K21= 0.8850 are the gains from the FIS it must be defuzzified from fuzzy value to
for the coupled tank system. Transfer function of the crisp value center of gravity rule method is used for
coupled tank model is as below. defuzzification.
The rules shown in the rule matrix are expressed as
shown below.

The analysis of this mathematical model is done using IF the level is low AND rate is zero THEN open fast valve.
MATLAB/Simulink toolbox for P, PI, and PID controllers.
Table-1: rule matrix
3. CONTROLLER DESIGN FOR COUPLED TANK SYSTEM
Level Low Okay High
3.1 PID controller design for the system

A simulated model for PID control system is as

shown fig.3 the difference of set point and measured Rate
variable goes to the controller. PID controller combines
the control action of Proportional, Integral and derivative Negative OF OS CF
controller where Proportional controller reduces the rise
time, integral controller eliminates steady state error and Zero OF NC CF
derivative controller reduces overshoots of the system.
Positive OF CS CF
PID controller involves three tuning parameters Kp, Ki and
Kd. In this paper parameters are tuned within
MATLAB/Simulink block signal constraint. This tool helps Where
in optimization of parameters very fast. In this tool we OF – open fast
used 0 to infinity tolerance for finding optimal solution or OS – open slowly
feasible solution. Parameters values found for P, PI and CF – close fast
PID controller which is shown below. CS – close slowly
Proportional controller- Kp=328.34 not found any feasible NC – no change
solution
Proportional plus Integral controller- Kp = 67.4043, 4. SIMULINK RESULTS AND COMPARISON
Ki=1.1154 found a feasible solution
Proportional plus Integral plus Derivative controller- The simulink model for coupled tank, PID
Kp=67.3394, Ki=1.1552, Kd=10.7037 found a feasible controller and fuzzy logic control system and their
solution simulated response to find rise time, setting time
This optimization method gives fast response. overshoot and steady state error of the controllers are
shown below.
3.2 Fuzzy logic controller design for the system

FLC consist of fuzzification, fuzzy rules and

defuzzification. Where fuzzification converts the crisp
inputs into fuzzy inputs. We used rate and level as an input
and valve as an output for this application. Fuzzification
next step is selection of membership function for input
and output variables. Triangular membership function is
used for both input and output variables. In the next step
range of subset for variables is selected within the display
range variables. Here we have used low, high and ok
subsets for error variable and negative, zero, positive
subsets for rate of change variable and OF, OS, CF, CS, NC
subsets for valve variable. Ranges of these subsets are
selected within the variables membership function range.
After the selection of range for each subset these subsets
are used for the designing of rules in rule editor tool box.
Response of the controller depends upon the rules. Nine
rules are designed for this application which is shown in
matrix form. Response of the designed rules can be viewed Fig.2: Simulink model for coupled tank

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1175
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072

Fig.3: Simulink model for PID control system

Fig.7: Simulink response of P controller

Fig.4: simulink model for fuzzy logic control system

Fig.8: simulink response of fuzzy logic controller

Fig.5: Simulink response of PID controller

Fig.9: Comparative Simulink response of P, PI, PID and

fuzzy logic controller for parameters
Fig.6: Simulink response of PI controller

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1176
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072

Table-2 Comparative evaluation of controllers in tabular [6] Sayed Kamaleddin Mousavi Mashidi, "Design fuzzy
form controller for synthesis water level", Journal of
Mathematics and Computer Science, pp 300-313, 2014.
Controll Rise Steady Peak Setting time
[7] S.Nithya, Abhay Singh Gour, "Intelligent Controller
ers time state oversho
Implementation in Real time for a non linear process",
error ot
IEEE transaction on process controller, pp 2508-2513,
2008.
P 1.9sec 0.9% 50% 20sec
[8] Meng Quinsong, Wang Qi, Wei Hongling, "Design of
fuzzy controller for liquid level control system based on
PI 3.41se 1.6% 18% 16sec MATLAB/RTW", proceedings of the 2 nd IEEE International
c conference on measurement, information and control,
Harbin, China, vol.2, pp 1090-1094, 2013.
PID 3.56se 1.6% 13% 12sec [9] Lian Li, Wengkuan Ding, "Optimization control strategy
c of Boiler water level based on fuzzy PID", proceedings of
the 28th IEEE Chinese control and decision conference,
China, pp 5893-5896, 2016.
FLC 3.25se 0 0 5sec
c

5 CONCLUSIONS

In this paper, we developed the mathematical

model of coupled tank system and design a simulink
model for coupled tank, PID controller and fuzzy logic
control system. Designed model are simulated within
MATLAB/Simulink and comparatively analyzed in terms
of rise time, steady state error, peak overshoot and setting
time. From the analysis we concluded that coupled tank
system with fuzzy logic controller gives better
performance. The simulated results shows that fuzzy logic
controller is a better option for real world applications as
we can see in the results shown in table 2.

REFERENCES

[1] S.Shamily, Parveena and Bhuvaneswar, "Intelligent

control and Adaptive control for Interacting system",
proceedings of IEEE International conference on
technological innovations in ICT for agriculture and rural
development, pp 116-121, 2015.
[2] Qu Qingly, Zhao Pingrong, Wu Yangxi, "Control system
of the Boiler water level based on Normalized PID",
proceedings of the 3rd IEEE International conference on
computer science and network technology, 1251-1254,
2013.
[3] S.A.Mali, C.B Kadu, B.J Parvat, "Design of fuzzy adaptive
PI controller for SISO system", proceedings of IEEE
International conference on energy system and
applications, pp 234-238, 2015.
[4] Michail Petrov, Ganchev Albena Taneva, "Fuzzy PID
control of nonlinear plants", 1st international IEEE
symposium on intelligent system, vol.1, pp 30-35, 2002.
[5] Muhammad Abid, "Fuzzy logic control of coupled liquid
tank system", IEEE transaction on process control, pp144-
147, 2005.

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1177