International Journal of Innovative Technology and Exploring Engineering (IJITEE)

ISSN: 2278-3075, Volume-1, Issue-1, June 2012

Comparison between Conventional PID and

Fuzzy Logic Controller for Liquid Flow Control:
Performance Evaluation of Fuzzy Logic and PID
Controller by Using MATLAB/Simulink
Gaurav, Amrit Kaur

Abstract:- Measuring the flow of liquids is a critical

need in many industrial plants. In recent years, flow
control has become a highly multi-disciplinary research
activity encompassing theoretical, computational and
experimental fluid dynamics. Fuzzy control is based on
fuzzy logic-a logical system that is much closer in spirit to
human thinking and natural language than traditional
logical systems. During the past several years, fuzzy
control has emerged as one of the most active and fruitful
areas for research in the applications of fuzzy set theory, Figure 1: A simple fuzzy system
especially in the realm of industrial processes, which do A simple fuzzy system consists of four blocks: A Fuzzifier,
not lend themselves to control by conventional methods Defuzzifier, inference engine and fuzzy rule knowledge
because of a lack of quantitative data regarding the input- base. Fuzzy Logic Controller (FLC) is an attractive choice
output relations. The fuzzy logic controller (FLC) based on when precise mathematical formulations are not possible
fuzzy logic provides a means of converting a linguistic [3]. Other advantages are:
control strategy based on expert knowledge into an
 It can work with less precise inputs.
automatic control strategy. Fuzzy Logic controller has
better stability, small overshoot, and fast response. In this
 It doesn‟t need fast processors.
Paper, performance analysis of the conventional PID
controller and fuzzy logic controller has been done by the It is more robust than other non-linear controllers.
use of Matlab and Simulink and in the end comparison of
various time domain parameters is done to prove that the II. Fuzzy Logic Controller Are Better Than
fuzzy logic controller has small overshoot and fast Conventional Controllers. WHY?
response as compared to PID controller. Fuzzy control has emerged one of the most active and
fruitful areas of research especially in industrial processes
Key Words:- Flow control, Conventional control, Fuzzy which do not rely upon the conventional methods because of
logic control, Simulink. lack of quantitative data regarding the input and output
relations. Fuzzy control is based on fuzzy logic, a logical
I. Introduction to Fuzzy Logic system which is much closer to human thinking and natural
The Fuzzy Logic tool was introduced by Lotfi Zadeh language than traditional logical systems [1]. Fuzzy logic
(1965), and is a mathematical tool for dealing with controller (FLC) based on fuzzy logic provides a means of
uncertainty. It offers to a soft computing partnership the converting a linguistic control strategy based on expert
important concept of computing with words. It provides a knowledge into an automatic control strategy. Fuzzification,
technique to deal with imprecision. The fuzzy theory defuzzification strategies and fuzzy control rules are used in
provides a mechanism for representing linguistic constructs fuzzy reasoning mechanism [2].
such as “many,” “low,” “medium,” “often,” “few.” In
general, the fuzzy logic provides an inference structure that Fuzzy logic control has been successfully used in various
enables appropriate human reasoning capabilities. Fuzzy application areas ranging from automatic train operation to
logic systems are suitable for approximate reasoning. Fuzzy flight systems. Fuzzy logic enables control engineers to
logic systems have faster and smoother response than efficiently develop control strategies in application areas
conventional systems and control complexity is less [3].The marked by low order dynamics with weak non linearities.
basic building blocks of a Fuzzy system given below.
Kamal et al. [4] concluded the Fuzzy logic control of
refrigerant flow. Refrigerant is the medium used to transfer
heat from one place to another in refrigeration and air-
Manuscript received on May 29, 2012. conditioning systems. Control of refrigerant flow in
Gaurav,*Student, **Assistant professor, University College of engineering,
Punjabi university, Patiala, India refrigeration and air-conditioning systems is essential to
Amrit Kaur,*Student, **Assistant professor,University College of improve their performance and to prolong their life. The
engineering,, Punjabi university, Patiala, India. performance of the fuzzy logic controller is compared with a

84
 Comparison between Conventional PID and Fuzzy Logic Controller for Liquid Flow Control: Performance
Evaluation of Fuzzy Logic and PID Controller by Using MATLAB/Simulink

well known existing commercial controller and it shows that when applied to systems, which are nonlinear, as controller
the fuzzy logic controller has achieved better performance tuning is difficult due to insufficient knowledge of the
[3]. Fuzzy logic control is able to handle imprecision and parameters of the system. Fluid flow system is a typical
uncertainty. example.Neuro fuzzy controller gave a better performance
compared to the PID controller. It gives better performance
Fahid et al. [4] concluded that Proportional integrated with reduced oscillations and faster settling time. The
Derivative (PID) controllers are widely used in process controller performance can still be improved by training the
control applications, but they exhibit the poor performance neural network with more number of input and output
when applied to systems, which are nonlinear, as controller combinations [6].
tuning is difficult due to insufficient knowledge of the
parameters of the system. Fluid flow system is a typical Depending on all the above findings and research going on
example.Neuro fuzzy controller gave a better performance in this field it has been found that flow control has become a
compared to the PID controller. It gives better performance highly multi-disciplinary research activity encompassing
with reduced oscillations and faster settling time [5]. The theoretical, computational and experimental fluid dynamics.
controller performance can still be improved by training the It is an emerging field having potential benefits in
neural network with more number of input and output aerodynamics and bio-medical engineering. So it was
combinations. proposed that Fuzzy logic flow controller can be
implemented.
Elangeshwaran et al. [6] Overall, fuzzy logic controller is a
good alternative to a PID controller, for flow measurement A) Objectives:
and control applications. From all the above discussions we
can conclude that Fuzzy Logic controller has better stability, To study flow measurement Techniques.
small overshoot, and fast response.
 To study flow control system using conventional controllers.
Overall, fuzzy logic controller is a good alternative to a PID
controller, for flow measurement and control applications. To study the fuzzy logic system.
From all the above discussions we can conclude that Fuzzy
Logic controller has better stability, small overshoot, and Design and development for flow control. Proper Fuzzy
fast response. logic can be implemented for the flow control of fluids.

III Problem Formulation:  To analyze the controller performance of conventional PID

A.S. Kamal et al. [3] (1996) set out to apply the fuzzy logic controller and Fuzzy logic controller.
to control the refrigerant flow of a refrigeration system.
Fuzzy logic is relatively easy to design and implement. Its IV Results and Discussions:
performance has been compared with that of a well-known Fuzzy control has emerged one of the most active and
commercial controller. Fuzzy logic achieves better control fruitful areas of research especially in industrial processes
and improves the performance of the system. The surprising which do not rely upon the conventional methods because of
poor performance of the commercial controller could be due lack of quantitative data regarding the input and output
to inappropriate setting of the controller gains which cannot relations. Fuzzy control is based on fuzzy logic, a logical
be adjusted by users. Fuzzy logic is a viable alternative to system which is much closer to human thinking and natural
conventional forms Fuzzy logic control is able to handle language than traditional logical systems. Fuzzy logic rules
imprecision and uncertainty [3]. are simple and do not require precise control algorithm.
Fuzzy logic systems are suitable for approximate reasoning.
Elangeshwaran et al. [9] (2006) illustrates the advantages of
a fuzzy based controller over a PID controller are derived Fuzzy Logic Controller (FLC) is an attractive choice when
from the experiment results. Better control performance, precise mathematical formulations are not possible. Other
robustness and overall stability can be expected from the advantages are:
fuzzy controller. Fundamentally, the Fuzzy concept is
merely a representation of the human cognitive and decision It can work with less precise inputs and more robust in
making process hence developing and tuning of the FIS is nature.
more intuitive than the PID controller [9].
 It doesn‟t need fast processors.
M.M et al. [12] (2011) concluded that Fuzzy control
combined with conventional PID controller constitutes an A) Liquid Flow Control System Design With Fuzzy logic
intelligent control, which adjusts the control parameters controller:
depending upon the error. A two input and three output
Fuzzy Logic Control (FLC) has excelled in dealing with
fuzzy adaptive PID control was presented by them. The
systems that are complex, ill-defined, non-linear or time-
controller was simulated in MATLAB environment. The
varying. FLC is relatively easy to implement, as it usually
simulation results show that the fuzzy adaptive PID
needs no mathematical model of the control system. Fuzzy
controller have better stability, small overshoot, fast
logic has rapidly become one of the most successful of
response [12].
today's technologies for developing sophisticated control
Fahid et al. [6](2002) concluded that Proportional integrated systems. The reason for which is very simple. It fills an
Derivative (PID) controllers are widely used in process important gap in engineering design methods left vacant by
control applications, but they exhibit the poor performance purely mathematical approaches (e.g. linear control design),

85
 International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-3075, Volume-1, Issue-1, June 2012

and purely logic-based approaches (e.g. expert systems) in editor, where the number of membership functions, and type
system design. Figure 2 shows the system architecture with of membership function is chose, such as trapezoidal,
Fuzzy logic controller. triangular, and Gaussian according to the process parameter.
The fuzzy logical operation is Fuzzification. For the
B) Simulink Results: computation to be relatively simple, the research use
triangular shape.
Simulink® is an environment for multi domain simulation
and Model-Based Design for dynamic and embedded
systems [3].

In this model we can see that Fuzzy logic controller and PID
has been used. The difference between them can be seen
from change scope.

Figure 2: Development and implementation of

liquidflows in MATLAB/Simulink Environme

In scope 1 the yellow line shows input and pink line shows
output and it can be seen that output is very close to input.
Figure 5: shows the changed scope of controller.

Figure 3: This scope shows the comparative analysis of Figure 6: Shows Number of inputs (Level & Rate) and
PID and Fuzzy logic controller. output (Liquid Flow) for designing Fuzzy Inference
In change scope we are analyzing the controller Structure (FIS) for Fuzzy Logic controller.
performance of conventional PID controller and FUZZY
LOGIC CONTROLLER respectively.

Figure 4: shows the change in scope of conventional

controller.
Figure 7: Membership function editor for 2 I/P’s (Level
The fuzzy set is defined by a function that maps objects in a & Rate) whose MF is Gaussian and O/P (Valve) whose
domain of concern to their membership value in the set. MF is triangular.
Such a function is called membership function and is usually
denoted by Greek symbol “μ”. Figure 6, shows the selection
of number of inputs and outputs in the form of membership
functions in order to design FIS. So, it resembles the
selection of two inputs – level, rate, and one output –liquid
flow valve. Figure 7, shows the Fuzzy Membership function
86
 Comparison between Conventional PID and Fuzzy Logic Controller for Liquid Flow Control: Performance
Evaluation of Fuzzy Logic and PID Controller by Using MATLAB/Simulink

Figure 11: This graph is plotted b/t amplitude and time

using Fuzzy shows step response

COMPARING VARIOUS TIME DOMAIN

SPECIFICATIONS:

Rise Settli %Ste

S Delay Transi
Time ng Peak ady
. Controller Time( ent
(Tr) Time Overshoot State
n Used Td)in Behav
in (Ts)i (Mp) in % Error
o sec ior
sec n sec (Ess)

Figure 8: Rule Viewer for Fuzzy controller

PID
Oscill
1. CONTRO 0.1 0.3 16 13 0
atory
LLER

2. FUZZY 1 1.03 3.85 2.7 Smoot -6

LOGIC h
CONTRO
LLER

There are basically two Fuzzification methods namely,

Mamdani and Sugeno, and generally used Defuzzification
methods are:
 Adaptive integration,
 Center of area,
 Center of gravity,
 Fuzzy clustering Defuzzification,
 First of maximum,
 Last of maximum,
Figure 9: Fig shows the corresponding surface viewer  Mean of maxima,
 Semi-linear Defuzzification, and
 Centroid method [6].

Basically Defuzzification is the process of converting the

fuzzy conclusion into the crisp one and above is the
different methods of defuzzification used.

D) Basic Steps of Fuzzy Inference Structure (FIS):

The algorithm of fuzzy rule-based inference consists of four

basic steps given as follows:

Fuzzy Matching: - Calculate the degree to which the input

Figure 10: This graph is plotted b/t amplitude and time data match the condition of the fuzzy rules.
using Conventional PID shows step response.
Inference: - Calculate the rule‟s conclusion based on its
matching degree.

Combination: - Combine the conclusion inferred by all

fuzzy rules into a final conclusion.

Defuzzification: - For applications that need a crisp output

(e.g., in control systems), this step is used to convert a fuzzy
conclusion into a crisp one [3].

E) Fuzzy Rules for Developing FIS:

Human beings make decisions based on rules. Although, we

may not be aware of it, all the decisions we make are all
based on computer like if-then statements.

87
 International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-3075, Volume-1, Issue-1, June 2012

IF – THEN FUZZY STATEMENTS FOR FUZZY REFERENCES

INFERENCE STRUCTURE (FIS) [1]. Rem Langari,”Past, present and future of fuzzy control:
A case for application of fuzzy logic in hierarchical
Fuzzy machines always tend to mimic the behaviour of man control,”IEEE, pp.760-765, 1999.
.Fuzzy rules also operate using a series of if-then statements. [2].Chuen Chien Lee, “Fuzzy logic in control systems i.e.
The fuzzy control rule is based on fuzzy decision-making, fuzzy logic controller,”IEEE Transactions on Systems,
which satisfies some input conditions and has an output man and cybernetics, Vol 20, No.2, March/April 1990.
result. [3]. J.Y.M. Cheung, A.S. Kamal,” Fuzzy Logic Control of
refrigerant flow”, UKACC International Conference on
V Simulation Results: Control „96, Conference Publication No. 427, 2-5
The Figure 10, 11 shows the formation of response of the September 1996.
system when using PID & Fuzzy Logic controllers [4]. Rajanbabu.N, Sreenadhan.S, Fahid K.A, Mohandas K
respectively. Fuzzy logic controller is used in this process P,”Design and implementation of a neuro-fuzzy
because of following reasons: controller for a flow system”, presentation at FAE
symposium, European university of lefke, Nov 2002.
 It can work with less precise inputs. [5]. A. Abdelgawadt, Adam Lewist, M. Elgamelt, Fadi Issa,
N.F. Tzeng , and M. Bayoumit,” Remote Measuring of
 It doesn‟t need fast processors. Flow Meters for Petroleum Engineering and Other
Industrial Applications” The International Workshop on
 It is more robust than other non-linear controllers. Computer Architecture for Machine Perception and
Sensing, pp.99-103, September 2007.
 Fuzzy controllers have better stability, small overshoot, [6]. Elangeshwaran Pathmanathan, Rosdiazli Ibrahim,”
and fast response. Development and Implementation of Fuzzy Logic
Controller for Flow Control Application,” Intelligent
After comparing the graphs of conventional PID and fuzzy
and Advanced Systems (ICIAS), International
logic controller as shown in figure10, 11 it is clear that
Conference on Digital Object Identifier, pp.1-6, 2010.
fuzzy logic has small overshoot and is having the fast
[7]. R. Manoj Manjunath,S, S. Janaki Raman ,” Fuzzy
response as compared to conventional PID Controller. Then,
Adaptive PID for Flow Control System based on OPC,”
various time domain specifications of both the controllers
IJCA Special Issue on “Computational Science - New
are compared such as:
Dimensions & Perspectives NCCSE”, 2011.
[8].http://en.wikipedia.org/wiki/Flow_meter.
 Rise Time(Tr) and Delay Time (Td),
[9].http://www.virtuallaboratories.com/html/fluid_a.htm
 Settling Time (Ts
 Peak Overshoot (Mp)
 Steady State Error (Ess)
 Transient Behaviour

VI Conclusion:
Overall the project‟s feasibility lies in the simplicity of its
implementation. The advantages of a fuzzy based controller
over a PID controller are derived from results. Better control
performance, robustness and overall stability can be
expected from the fuzzy controller. Fuzzy controllers have
better stability, small overshoot, and fast response.From the
results the following parameters can be observed. Hence, Mr. Gaurav is currently pursuing M.TECH (final year) in department of
fuzzy logic controller is introduced for controlling fluid Electronics and Communication Engineering at University College of
flows. Engineering, Punjabi University, Patiala. He has completed his B.TECH in
trade electronics and communication engineering from University College
of Engineering, Punjabi University, Patiala. His areas of interest are Fuzzy
1) Even though, the PID controller produces the response logic and Neural Networks, Micro Electronics.
with lower delay time and rise time compared with fuzzy
logic controller, but it offers very high settling time due to
the oscillatory behaviour in transient period. It has severe
oscillations with a very high peak overshoot of 13% which
causes the damage in the system performance.
2) The proposed Fuzzy logic controller can effectively
eliminate these dangerous oscillations and provides smooth
operation in transient period.
Hence, it is concluded that the conventional PID controller
could not be used for the control of non-linear processes like
fluid flows. So, the proposed fuzzy logic based controller
design can be a preferable choice for this.

88