Fuzzy Logic Modelling, Simulation and Control: A Review: Rahul Malhotra, Narinder Singh, Yaduvir Singh
Fuzzy Logic Modelling, Simulation and Control: A Review: Rahul Malhotra, Narinder Singh, Yaduvir Singh
Fuzzy Logic Modelling, Simulation and Control: A Review: Rahul Malhotra, Narinder Singh, Yaduvir Singh
C. Fuzzification
Fuzzification is a process in which the inputs are classified
to corresponding linguistic terms to get degree of fulfillment
(DOF). It means adding uncertainty by design to crisp sets or
to sets that are already fuzzy. This process allows converting a
crisp value into a fuzzy one with the two common ways namely
fuzzy singleton and fuzzy set
Aui is a fuzzy singleton, such that
1 if u = u i
µAui (u) =
0 otherwise …(5)
Aui is a fuzzy set such that
1 if u = u i
µAui (u) =
decreases.from.1.as.u.moves.from.u i …(6) Fig. 2: Fuzzification and defuzzification methods
Singleton method of fuzzification represented in equation 3.31
is used in implementation where there is no noise. Fig. 3.7 E. Fuzzy Modeling
shows the fuzzification performed by singleton method. A fuzzy model uses fuzzy rules, which are linguistic statements
involving fuzzy sets, fuzzy logic and fuzzy inference. Fuzzy
rule play a key role in representing expert control, modeling
knowledge and experience in linking the input variables of
fuzzy controllers to output variables. Two major types of rules,
viz. Mamdani fuzzy rules and Takagi-Sugeno are available for
modeling a fuzzy rule base.
1. Mamdani Model
Fuzzy logic control attempts to design the informal nature of
the control design process. The Mamdani architecture is the
way to design a fuzzy control system. The Mamdani architecture
for fuzzy logic control is proposed by E. H. Mamdani in 1974,
Fig. 1: Singleton fuzzification is that, in the absence of an explicit plant model and-or clear
statement of control design objectives, informal knowledge
D. Defuzzification of the operation of the given plant can be coded in terms of
Fuzzification Defuzzification is the inverse process of fuzzification, IF-THEN or condition-action, rules and form the basis for the
in which, the output linguistic terms are converted into crisp linguistic control strategy, for example, a fuzzy rule
values according to their degree of fulfillment (DOF). There IF speed is high AND acceleration is small THEN braking is
are many techniques to perform defuzzification operation. The modest …(9)
choice of defuzzification method is fairly technical matter and is where the speed and acceleration are input variables and on
strictly as per the requirements of the problem. The important the other hand, braking is an output variable. High, small and
methods to perform defuzzification are max-membership, modest are fuzzy sets, and the first two sets are input fuzzy
centre of gravity method, weighted average method, mean- sets and the last one is the output fuzzy set.
max method and centre of sums method.
The process of fuzzy aggregation and fuzzy defuzzification is 2. Takagi-Sugeno model
explained by the following rules. Takagi-Sugeno model is an alternative architecture introduced
IF x1 is small AND x2 is medium THEN y is small …(7) by Takagi and Sugeno, uses a combination of linguistic rules
IF x1 is medium AND x2 is big THEN y is big …(8) and linear functions to form a fuzzy logic control strategy. In the
Fig. 2 shows the process of fuzzification, aggregation and Takagi-Sugeno fuzzy system, the conclusion of the fuzzy set is
defuzzificaiton operation together for a process control system. not a fuzzy set but a crisp function of the inputs. Fuzzy IF-THEN
Centre of sums method of defuzzification is used to implement rule can be also designed using Takagi-Sugeno modeling and
the two rules as given in equations 7 and 8. Generally, any can be expressed as
process control system sequentially performs the above
operations to control a plant. IF (x1 is A1) AND (x2 is B1) THEN y = f1(x) = c10+c11x1+ c12x2
…(10)
The functions f(x) of the input vector x = [x1, x2, ..] can be very
complex but the most commonly used function is linear function
that can be described by the coefficients [c0, c1, c2, ..]. This
architecture has gained support because of its efficiency and
will most likely be the architecture for many future applications
184 International Journal of Computer Science and Technology w w w. i j c s t. c o m
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online) IJCST Vol. 1, Issue 2, December 2010
of fuzzy logic control. is to create the control actions, in fuzzy terms, according to
the information provided by the fuzzification module. The
F. DESIGN OF FUZZY CONTROL SYSTEM defuzzification module is the connection between the control
The task of modeling complex real-world processes for the rule base and the physical plant to be controlled, which
control system design is a challenging engineering problem. plays the role of transformer mapping the controller outputs.
Even if a relatively accurate model to control the process Defuzzification creates a crisp, overall control signal, by
is developed, it is often too complex to use it in controller combining all possible control outputs from the rule base into
development, as much simpler process model is required by a weighted average notation. Input and output filters are used
most of the conventional control design techniques. Fuzzy for signal preprocessing. A typical feedback control system with
logic control was first introduced by Mamdani and is based direct fuzzy controller is shown in Fig. 4.
on Zade’s theory of fuzzy sets. A systematic design procedure
is to be followed while designing a complex control system.
With an intuitive understanding of the process and establishing
control design objectives, the process can be controlled by the
following steps:
Step I. Developing a model of the process dynamics.
Step II. Designing a controller using its mathematical
model.
Step III. Simulation based analysis to study the performance
of the control system. Fig. 4: Direct fuzzy logic controller with feedback system
Step IV. Implementing the controller and evaluating its Consider the above plant controlling the complex process using
performance on the plant. fuzzy logic controller. To implement the fuzzy rules on a digital
Conventional controllers viz. proportional integral derivatives computer, the discrete-time version is used. Another alternate
(PID) employ a process modeled, by differential equations. for determining the weight of control is to design fuzzy rule
On the other hand, fuzzy control focuses on gaining intuitive base in tabular form for the complex control system, that is,
understanding for better control of the process, and this representation of linguistic terms in negative big (NB), negative
information is then loaded directly into fuzzy controller. The medium (NM), negative small (NS), zero (ZO), positive small (PS),
performance objectives and the design constraints usually positive medium (PM) and positive big (PB). A typical rule base
involve the stability, rise time, overshoot, settling time and in tabular form to control the process is shown in Table 1.
steady state error of process controllers. The conventional TABLE I
control techniques utilize mathematical models for controlling IF-THEN RULE BASE FOR FUZZY LOGIC CONTROL
the process. Some heuristics do not fit into mathematical e(t)
framework, which compelled the modern control system to u(t)
use soft computing techniques for improving the efficiency NB NM NS ZO PS PM PB
and efficacy of the process to be controlled. NB NB NB NB NB NM NS ZO
NM NB NB NB NM NS ZO PS
G. Direct Fuzzy control NS NB NB NM NS ZO PS PM
Fuzzy logic is an infinite-valued logic and the truth of proposition ∆e(t) ZO NB NM NS ZO PS PM PB
in such a framework can take any value in the unit interval. Today PS NM NS ZO PS PM PB PB
fuzzy logic has become a key word of marketing in consumer PM NS ZO PS PM PB PB PB
products like washing machine, camcorders, televisison and PB ZO PS PM PB PB PB PB
air conditioners. The general structure of fuzzy logic controller
consist of three basic portions, that is, fuzzification unit at the Using above table, fuzzy logic rule base can be designed for the
input terminal, the inference engine built on fuzzy logic control different processes to be controlled. Let us take an example
rule base in the core, and the defuzzification unit at the output of the rule.
terminal as shown in Fig. 3. IF ∆e(t) is NS AND u(t) is NB THEN e(t) is NB. …(12)
knowledge about the structure of the plant. The main idea is controller design. Another uses design of fuzzy system with
to use the error signal, which is the difference between the precise parameter uncertainty. The design of fuzzy system
reference signal and the system output, to drive the fuzzy guarantees the stability of the system. Simulation shows the
logic controller, so as to create a new control action. These designed fuzzy controller smoothly achieves the backing up
control actions should be able to alter the system outputs in control of the truck-trailor from all initial positions.
such a way that the error signal is reduced until satisfactory The use of fuzzy logic in telecom system is suggested by Ghosh
performance is achieved. Successful design of fuzzy logic et.al (1998). The work discussed about the dynamic network
controller, further needs to optimize the fuzzy rule base and problems and reviews the research efforts in queuing, buffer
an error. The design of fuzzy sets and rule base has been management, congestion, mitigation, call acceptance, network
automated by the use of genetic algorithms (GAs) which are management and quantitative performance evaluation in
powerful search techniques. networks.
Extensive research has been done to develop fuzzy logic models A systematic neural-fuzzy modeling framework that includes
successfully for the different application domains. Many of the initial fuzzy model self-generation, input selection, partition,
the researchers, though, have faced the dilemma that much validation and parameter optimization was proposed by Min-
of their data is plagued by an uncertainty, vagueness and Yon Chen et.al (2001). The proposed approach was used for
approximations. This review work on fuzzy logic elaborates non-linear system identification and mechanical property
the scope of fuzzy logic implementation in process control prediction in hot-rolled steels.
applications. A permanent magnet synchronous servo motor drive using
Yang et.al (2004) discussed a robust adaptive control problem integral proportional controller and a online trained fuzzy neural
for a general class of strict feedback uncertain non-linear network controller was implemented by Faa-Jeng et.al (1998).
systems. Takagi-Sukeno method of fuzzy logic is used to The membership function and connective weights are online
approximate the uncertainties. The case studies of pendulum trained according to the model following error of the states.
system of motor, a one-line robot, a strip roll stabilization Fillippetti F. et. al (1998) discussed artificial intelligence
with actuator and a single-link manipulator demonstrates the techniques in induction motor diagnosis, including the speed
effectiveness and performance of proposed schemes. ripple effect. A new procedure based on model is developed
Fuzzy logic is applied to control the speed of an induction motor which leads to a new diagnostic index, independent of the
by Ibrahim et.al (2002). The comparison of fuzzy logic with machine operating condition and inertia value that allows
conventional proportional intergral controller is made and the the implementation of the diagnostic system with minimum
superiority of fuzzy logic method is quoted in the study. configuration intelligence.
A methodology for designing and tuning the scaling gains of Lin CM and Hiyana (1991) implemented control strategy that
conventional fuzzy logic controller was discussed and the new incorporates proportional integral controller with simple fuzzy
concept of fuzzy transfer function was invented to connect logic controller. In the proposed strategy, proportional integral
these gains of the well tuned conventional fuzzy logic controller. controller is used to ensure fast transient response and zero
Furthermore, the relationship between the scaling gains and steady state error for step inputs or end point control, whereas
performance can be deduced to produce the comparative tuning fuzzy logic is used to enhance the damping characteristics of
algorithm, which can tune the scaling gains to their optimum by overall system. The effectiveness of fuzzy logic controller is
less trial and error. The utility of fuzzy system is demonstrated described through simulations.
by emphasizing analog mode hardware by Yamakawa (1993). Molo Fu et.al (1996) discussed how phase locked loop (PLL)
The difference between deterministic word and fuzzy word is can be incorporated with fuzzy logic to control the speed of
discussed and the system was mathematically modeled and induction motor. The proposed system takes the advantages of
fuzzy logic is applied to stabilize the inverted pendulum. phase locked loop (PLL) for speed regulation and advantages
Palaniappan et.al (2002) proposed a new brain computer like intuitiveness, simplicity, easy implementation and minimal
interface using fuzzy ARTMAP with an objective to classify the knowledge of the system dynamics of fuzzy logic to evaluate
best three of five mental tasks for each subject using power the performance of induction motor and concluded that the
spectral density of electroencephalogram (EEG) signals. The combination of phase locked loop (PLL) and fuzzy logic (FL) can
morse code scheme was used to translate the signal of new achieve precise speed control with fast response.
brain computer interface using fuzzy ARTMAP into english The general methodology to fuzzy logic to fuse the available
letters which could be benefited as communicator method for information for spatial load forecasting was discussed by Mo-
paralyzed parents. Yuen Chow et.al (1996). The proposed scheme can provide
Mudi R.K. and Pal (1999) proposed a robust self tuning scheme distribution planners other alternatives to aggregate their
for fuzzy logic control. The output scaling factor is adjusted information for spatial load forecasting.
online by fuzzy rules according to current trend of the controlled Bowles and Pelaez et.al (1995) gives an idea about fuzzy
process. The rule base for tuning the output is defined as an error logic implementation to analyze the reliability of the system
and change in error of the controlled variable using most natural which includes subjective judgements, uncertain data and
and unbiased membership function. Several performance approximate system models.
measures such as peak, overshoot, settling time, rise time, Development and implementation of adaptive fuzzy neural
integral of absolute error (IAE), and integral of time and absolute controller for brushless drives was done by Rubaai et.al (2002).
error (ITAE) have been discussed which remarkably shows the The characteristic rules and their membership function of fuzzy
importance of fuzzy logic control technique over conventional systems are represented as the processing nodes of neural
control techniques. networks. The gradient decent learning algorithm is used for
A robust stabilization problem of fuzzy control systems and its tuning of fuzzy rules. The effectiveness of proposed scheme
application to backing up control of a truck-trailor was discussed is demonstrated by comparing it with proportional integral
by Tanaka K. et.al (1994). Two problems were considered in controller.
this work, First uses, non-robust controller design and robust The design and experimental verification of fuzzy control system
186 International Journal of Computer Science and Technology w w w. i j c s t. c o m
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online) IJCST Vol. 1, Issue 2, December 2010
applied to brushless DC motor, which shows the excellent very high flow they consume much power. The combination of
tracking performance of the proposed control system, and gas turbine compressor is highly reliable. Hence, the turbine
have convincingly demonstrated the usefulness of the hybrid compressor plays a significant role in the operation of the
fuzzy controller in high performance drives with uncertainties plants. High pressure steam (HPS) is used to drive the turbine.
by Rubaai A et.al (2001). The turbine is coupled with the compressor and eventually the
Field programmable gate array (FPGA) implementation of compressor is run by the gas turbine. The hydraulic governor
embedded fuzzy logic control for robotic applications was is used to control the flow of high pressure steam. It acts as
implemented by Sanchez et.al (2007). They discussed a control a control valve which is used to throttle the amount of steam
strategy that eases the implementation of embedded fuzzy going to turbine section.
control on programmable chips. The design methodology has Conventional controllers like proportional integral derivative
been applied to realization of a control system for solving the (PID) are employed in every facet of process control system.
navigation tasks of an autonomous vehicle. Proportional integral derivative (PID) controllers are used in
Fuzzy control for mechanical system is demonstrated by small industries as well as in high technology industries like
Kulczycki (2000) for design of sub optimal feedback controllers. refineries, ship buildings etc. Proportional integral derivative
This study deals with the time optimal control for mechanical (PID) controllers can be fine tuned by Ziegler-Nichols method
systems with a discontinuous and uncertain model of resistance described in the previous section.
to motion. Due to inherent disadvantages of proportional integral derivative
A survey on fuzzy logic has been carried out by Yen. J (1999). (PID) controller, fuzzy logic (FL) based controller is implemented
In this review, the history of fuzzy logic and the development to control the flow of high pressure steam.
of fuzzy rules viz. Takagi-Sukeno and Mamdani was discussed. The fuzzy logic controller (FLC) gives much lower overshoot,
Further, some challenges regarding automated identification settling time and peak time than the conventional proportional
of fuzzy rule base models have been explored. integral derivative (PID) controller. Typical value of overshoot
Some Experimental examples of the process controllers have is 8.75% and settling time and peak time are 6.8 seconds
been implemented using fuzzy logic technique and compared and 6.2 seconds respectively, which shows the comparative
with the conventional techniques. improvement in the fuzzy logic based controller than
conventional proportional integral derivative (PID) controller
G. DTC Induction Motor for different process control parameters.
Intelligent techniques are used to achieve high performance
decoupled flux and torque control. Induction motors are I. DC Servo Motor
highly non-linear and complex in nature. The speed control Servo mechanisms are receiving wide attention for industrial
of an induction motor is a challenging engineering problem. applications because of their high torque density, high efficiency
Intelligent controllers are used to minimize flux and torque and small size. They are also known as control motors. Due to
ripples. Direct torque control (DTC) strategy is applied to control multivariate and non-linear in nature, it is difficult to achieve
an induction motor using soft computing techniques. The the optimum speed of the motor. Conventional proportional
parameters of typical induction motor taken up for the study integral (PI) controllers suffer from their inherent losses. The
are given below. fuzzy control has been focus to control the speed of servo
TABLE 2: motor.
PARAMETERS FOR TYPICAL INDUCTION MOTOR TABLE 3:
Parameter Magnitude PARAMETERS FOR SPEED CONTROL
Symbol Parameter Magnitude
Power 1KW
Phase 3 Ra Resistance of armature 2.1975 Ω
Stator Resistance 6.2 ohms La Inductance of armature 0.0063 H
winding
Rotor Resistance 7.3 ohms Ea Applied Armature Voltage 240 V
Magnetic Inductance 0.6014 H Tm Torque developed by motor 3 Nm
Stator and Rotor Inductance 0.0391 H J Moment of inertia 0.0236 kgm-2
f0 Viscous friction coefficient 0.015 (Nm/
Nominal Speed 1300 rpm (rad/sec))
Nominal Torque 5.2 Nm Km Torque Constant 1.78
Moment of Inertia 0.006 kg-m2 N Rated Speed 1200 rpm
P Rated Power 120 W
It is inferred from the results, that fuzzy logic approach can reduce A maximum overshoot of 9.33 percent, settling time and peak
the flux and torque errors than conventional controllers. time of 1.8 seconds and 2.3 seconds were noted which are
comparably less than conventional proportional integral (PI)
H. Gas Turbine controller. Fuzzy logic controllers can cope with non-linearities,
Gas turbines are highly non-linear plants that have multiple load disturbances but tuning of fuzzy logic (FL) parameters and
inputs and outputs. The operating conditions span extremes stabilities are its main problems. If the control parameters are
of temperature, pressure and load conditions. In chemical fine tuned then the system will be able to control the speed
plants, gas turbine is used to drive the compressor system. A in a better manner.
turbine is a mechanical device that extracts thermal energy
from the pressurized steam, and converts it into rotary motion. III. Conclusion And Future Scope
These compressors are normally of centrifugal type. To handle The paper has presented a comprehensive review of fuzzy
w w w. i j c s t. c o m International Journal of Computer Science and Technology 187
IJCST Vol. 1, Issue 2, December 2010 ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online)
logic and its applications to control a plant for minimizing fused fuzzy systems via genetic algorithms,” Trans. Ind.
the errors. It is evident from the results of the different case Electron of IEEE, Vol. 49, no. 3, pp. 685-692, 2002.
studies taken up in this work that the fuzzy logic are a great [16] W.-S. Oh, Y.-T. Kim, C.-S. Kim, T.-S. Kown, and H.-J. Kim,
deal of research for tackling non-linear complex process control “Speed Control of Induction Motor Using Genetic Algorithm
systems. A conclusion can be drawn that the fuzzy logic is most Based Fuzzy Controller,” Industrial Electronics Society
robust and systematic approach for controlling the process. Conference of IEEE, vol. 3, no. 2, pp. 625-629, 1999.
Control of process can be better optimized with evolutionary [17] Wai Phyo Aung, “Analysis on Modeling and Simulink
techniques. of DC Motor and its Driving System Used for Wheeled
Mobile robot,” World Academy of Science, engineering
References and technology, 32, 2007, pp. 299-306.
[1] Atef Saleh Othman Al-Mashakbeh, “PID Control of DC [18] XiaChangliang, Guo Peijian, Shi Tingna and Wang
Motor,” European Journal of Scientific Research, Vol. 35, Mingchao, “Speed control of brushless DC motor using
No. 2, 2009, pp. 198-203. genetic algorithm based fuzzy controller”, Proceedings of
[2] B Allaoua, B Gasbaoui, B Mebarki, “Setting up PID DC the International Conference on Intelligent Mechatronics
motor speed control alteration parameters using PSO and Automation Chengdu,China August 2004, pp.460-
strategy,” Leonardo Electronic Journal of Practices and 464.
technologies, 2009, pp. 19-32. [19] Y.H. Bharathi, B.R. Rekha, P. Bhaskar, C.S. Parvathi and A.B.
[3] Boumediène ALLAOUA, Brahim GASBAOUI and Brahim Kulkarni, “Multi-input Fuzzy Logic Controller for Brushless
MEBARKI, “Setting Up PID DC Motor Speed Control dc Motor Drives”, Defence Science Journal, Vol. 58, No.
Alteration Parameters Using Particle Swarm Optimization 1, January 2008, pp. 147-158.
Strategy”, Leonardo Electronic Journal of Practices and [20] Y. F. Li, and C. C. Lau, “Development of fuzzy algorithms
Technologies, Issue 14, January-June 2009, pp. 19-32 for servo systems,” IEEE Control System Magazine, vol.
[4] Bhim Singh, B P Singh and (Ms) K Jain, “Implementation of 9, no. 3, pp. 65-72, 1989.
DSP Based Digital Speed Controller for Permanent Magnet
Brushless dc Motor”, IE(I) Journal-EL, Vol 84, June 2003,
pp.16-21.
[5] B Mrozek, Z Mrozek, “Modeling and Fuzzy Control of DC
Drive,” in Proceedings of 14th European Simulation Multi
conference ESM 2000, pp 186-190.
[6] C. C. Lee, “Fuzzy logic in control systems: fuzzy logic
controller, part- I”, IEEE Trans. on Systems, Man and
Cybernet, vol. 20, no. 2, pp. 404- 418, Mar./Apr. 1990.
[7] C.-K. Lee, and W.-H. Pang, “A Brushless DC Motor Speed
Control System Using Fuzzy Rules,” Power Electronics
and Variable-Speed Drives Conference, pp. 101-106, Oct
1994.
[8] Elke Laubwald, “Servo Control Systems: DC servo
mechanisms”, citation from control-systems-principles.
co.uk.
[9] G.C.D. Sousa, and B. K. Bose, “A fuzzy set theory based
control of a phase-controlled converter dc machine drive,”
IEEE Trans. on Industry Applications, vol. 30, no. 1, pp.
34-44, Jan/Feb 1994.
[10] G.Madhusudhana Rao, B.V.SankerRam, B.Smapath Kumar
and K.Vijay Kumar, “Speed Control of BLDC Motor Using
DSP”, International Journal of Engineering Science and
Technology, Vol. 2(3), 2010, pp.143-147.
[11] K. B. Mohanty, A. Routray, and N. K. De, “Design of a fuzzy
sliding mode controller for a field oriented induction motor
drive,” Journal of Systems Society of India, ‘Paritantra’,
vol. 6, no. 1, pp. 8-16, Aug. 2001.
[12] K. B. Mohanty, “Design and comparative study of sliding
mode and fuzzy sliding mode controllers for an induction
motor drive”, Procc. of 25th National Systems Conf., Dec.
2001, Coimbatore, pp 66-71.
[13] Mehmet Akar and Ismail Temiz, “Motion controller design
for the speed control of dc servo motor”, International
Journal of Applied Mathematics and Informatics, Issue
4, Volume 1, 2007, pp.131-137.
[14] P. K. Nandam, and P. C. Sen, “A comparative study of
proportional- integral (P-I) and integral-
proportional (I-P) controllers for dc motor drives,” Int. Jour.
of Control, vol. 44, pp. 283-297, 1986.
[15] P.-T. Chan, A.-B. Rad, and K.-M. Tsang, “Optimization of
188 International Journal of Computer Science and Technology w w w. i j c s t. c o m