©2010 International Journal of Computer Applications (0975 - 8887)

Volume 1 – No. 27

Design and Implementation of Adaptive Fuzzy Controller

for Speed Control of Brushless DC Motors
N.Senthil Kumar C.Senthil Kumar
Professor, EEE Department Assistant Professor, EEE Department,
Mepco schlenk Engineering College, PSR Engineering College, Sivakasi
Sivakasi, Tamilnadu India Tamilnadu, India

ABSTRACT
This paper presents the design and implementation of an adaptive model [5, 6]. The fuzzy logic controller uses fuzzy logic as a
fuzzy logic controller for the speed control of brushless dc motors. design methodology, which can be applied in developing non-
The proposed system uses an adaptation of the slope of the linear system for embedded control. Simplicity and less intensive
membership functions of the variables used in the conventional mathematical design requirements are the most important features
fuzzy controller based on the magnitude of the error. A simulation of the FLC.
analysis of the fuzzy controller and the adaptive fuzzy controller Fuzzy logic control is derived from fuzzy set theory introduced
are done and their performances are compared. Simulation results by Zadeh in 1965. In fuzzy set theory, the transition between
of both fuzzy and adaptive fuzzy controllers are presented. The membership and non-membership can be gradual. Therefore,
adaptive fuzzy controller is better than the fuzzy controller based boundaries of fuzzy sets can be vague and ambiguous, making it
on the performance parameters considered. An experimental useful for approximate systems. Fuzzy Logic controller is an
implementation of the designed adaptive fuzzy controller on an attractive choice when precise mathematical formulations are not
embedded microcontroller is also presented. possible [7, 8].
Keywords PROPOSED SYSTEM
Brushless DC motor, Fuzzy controller, Adaptive fuzzy controller,
microcontroller.

INTRODUCTION
Recently , permanent magnet brushless dc motor (PMBLDC) is
very popular because of its attractive features such as high
starting torque, high efficiency, low maintenance cost, absence of
mechanical commutator, high speed operation, low volume to
torque ratio, elimination of sparking and electromagnetic
disturbances, noise.
A PMBLDC motor is inside out construction of DC motor. The
efficiency is likely to be higher than DC motor of equal size and
the absence of commutator and brushes, reduces the motor length. Figure 1. Proposed system.
Hence the lateral stiffness of the motor is increased, allowing for Figure 1 shows the block diagram of proposed system. The
high speeds [1, 2]. The power electronic converters required in system above is composed of brushless dc motor, six step inverter,
brushless dc motor are similar in topology to the PWM inverter gate drive for inverter, fuzzy controller and switching logic. Due
used in induction motor drives. to the presence of parameter variation and load disturbance in a
Nowadays brushless dc motors are used in various applications BLDC motor, closed loop control is necessary, to obtain a
such as defense, industries, robotics, etc. In these applications, desirable behavior. BLDC motor has three phase windings on
motor should be precisely controlled so as to give desired stator and Permanent Magnet on rotor. In order to define the shaft
performance. The classical controller need accurate mathematical position, rotor position sensor is necessary. The sensor senses the
model of the system and can perform well only under linear rotor shaft position and signals. The processed signals are given to
condition. Since the PMBLDC motor is highly coupled non-linear the fuzzy controller. The output of the controller is used to
multivariable system, it is difficult to obtain its accurate provide switching signals for the inverter from which the speed of
mathematical model. Hence there is a need for intelligent the motor can be controlled.
controller. So an attempt is made to develop fuzzy controller for
PMBLDC motor.
FUZZY AS A CONTROL TOOL
The fuzzy logic controller (FLC) is indeed capable of Generally PI controller is widely used in BLDC motor control;
providing the high accuracy required by high performance drive however it does not give satisfactory results when control
system without the need of mathematical model [3, 4]. FLC parameters and loading condition changes rapidly [3]. The fuzzy
accommodates non-linearity without utilization of mathematical logic controller (FLC) will guarantee a stable operation, even if
there is a change in motor parameters and load disturbances. The

36
 ©2010 International Journal of Computer Applications (0975 - 8887)
Volume 1 – No. 27

reason is obvious; any control system maps the input space to the
output space. Generally, a desired set of outputs are calculated for
a given set of inputs. This mathematical calculation is represented
with a formula, which demonstrates the system behavior.
However, this mathematical formula may be too complex to use
for the real world issues. In these cases, fuzzy logic provides a
useful methodology to create a practical solution for controlling
complex systems. It is not necessary to know the exact model of
such complex systems in order to design a FLC. It is sufficient to
understand the general behavior of the system. Fuzzy logic
enables the designer to express the general behavior of the
systems in an easier (linguistic) manner where it is allowed to use
words and sentences instead of numbers and equations. This is
accomplished by forming IF-THEN rules which describe the
characteristics of the system. High degree of automation and
robust nonlinear control is also possible by means of fuzzy
controller.

DESIGN OF FUZZY CONTROLLER

The purpose of the speed control of a brushless dc motor is to

arrange the applied voltage in order to reach the reference speed.
An error is determined by the difference between the actual speed
and the reference speed. The applied voltage should be changed
by increasing or decreasing the duty cycle of power transistors in
order to minimize the error[4,5].In order to accomplish this task
fuzzy controller is designed. Error and change in error are the
inputs for the fuzzy controller whereas the output of the controller
is change in duty cycle. Two input single output Mamdani type of
fuzzy controller with 25 rules is designed for this work. Design of
fuzzy controller involves three steps namely fuzzification,
inference mechanism and defuzzification.

Fuzzification Fig.2 Membership functions used for fuzzy controller.

Fuzzy logic uses linguistic variables instead of numerical
variables. The process of converting a numerical variable in to a
linguistic variable is called fuzzification. Five linguistic variables
Negative Big(NB), Negative Small(NS), Zero(Z), Positive
4.2 The inference mechanism
Small(PS), Positive Big(PB) are used in this work. Triangular Table1: Rule base
membership function is assigned for input and output variables
defined in different universe of discourses. They are shown
CE E NB NS Z PS PB
below.

NB NB NB NB NS Z

NS NB NB NS Z PS

Z NB NS Z PS PB

PS NS Z PS PB PB

PB Z PS PB PB PB

37
 ©2010 International Journal of Computer Applications (0975 - 8887)
Volume 1 – No. 27

The rules are in the following format. If error is Ai, and

change in error is Bi then output is Ci. Here the if “part” of a rule
is called the rule-antecedent and is a description of a process state
in terms of a logical combination of atomic fuzzy propositions.
The “then” part of the rule is called the rule consequent and is a
description of the control output in terms of logical combinations
of fuzzy propositions. The rule table for the designed fuzzy
controller is given in the table 1.From the rule table the rules are
manipulated as follows [4,5].
Rule1: If error is NB, and change in error is NB then
output is NB
Rule2: If error is NB, and change in error is NS then
output is NB Fig.3 Membership function used for error in the range -1 to -
Rule3: If error is NB, and change in error is Z then 0.5 and 0.5 to 1
output is NB
.
.
.
Rule25: If error is PB, and change in error is PB then
output is PB.
4.3. Defuzzification
The reverse of fuzzification is called defuzzification. The use of
FLC produces required output in a linguistic variable. According
to real world requirements, the linguistic variables have to be
transformed to crisp Output. There are many methods of
defuzzification. Centroid method of defuzzification is used in this
work. The defuzzified output is obtained by the following Fig.4 Membership function used for error in the range -0.5 to
equation 0.5

∫ µ ( z ) zdz Computer Simulation

The closed loop model of the machine was developed and
Ż=
∫ µ ( z ) dz simulated with fuzzy controller. The simulation is carried by using
the software package MATLAB. The Simulink block diagram
used is shown in figure 5. Figure 6 shows the structure of the
fuzzy controller used. The Simulink model has the BLDC model,
Adaptive Fuzzy Controller which is fed from the inverter. Inverter receives the gate pulses
Adaptive fuzzy controller is one which provides provision for from the PWM Gate block. This block in turn receives two inputs
changing the parameters of fuzzy system based on performance - switching signal generated from encoder and duty cycle
index. Parameters of adaptation are generated from the fuzzy controller block. The step change in
o the scaling factors for each variable. speed reference and load is given at the pre-specified time. The
controller performance is tracked by observing the speed after the
o the fuzzy set representing the meaning of
input command.
linguistic variables.
o the if-then rules.[5]

In this work the slope of the membership function of error and

change in error are changed according to the values of error.
When the error lies in the range -1 to -0.5 and 0.5 to 1, the
membership function shown in figure 3 is used whereas when the
error lies in the range -0.5 to 0.5, the membership function shown
in figure 4 is used. This is accomplished in simulation by using
embedded Matlab function in simulink tool box. Fig.5 Simulink model of proposed system

38
 ©2010 International Journal of Computer Applications (0975 - 8887)
Volume 1 – No. 27

Fig.6 Structure of fuzzy controller used in this work

Fig.9 Simulated response of fuzzy controller for change in

Simulation Results speed with load change from 0 to 100% of the rated value at
The simulation results for step change in reference speed with the 1sec.
fuzzy controller are shown in figures 7, 8 and 9. Figure 7 shows
the simulated response for the step change in reference speed from
0 to 100% of the rated value at 0sec and 100% to 50% of the rated
value at 1sec. Figure 8 shows the simulated response for the step
change in reference speed from 0 to 50% of the rated value at 0sec
and 50% to 100% of the rated speed value at 1sec. Figure 9 shows
the simulated response for the step change in reference speed from
0 to 75% of the rated value for 2s with load torque changed from
0 to 100% of the rated value at 1sec. Simulation results for the
same conditions with adaptive fuzzy controller are shown in
figures 10, 11 and 12 respectively.

Fig.10 Simulated response of adaptive fuzzy controller for step

change in reference speed from rated value to 50% of the
rated value at 1sec

Fig.7 Simulated response of fuzzy controller for step change in

reference speed from rated value to 50% of the rated value at
1sec

Fig.11 Simulated response of adaptive fuzzy controller for step

change in reference speed from 50% of the rated value to the
rated value at 1sec

Fig.8 Simulated response of fuzzy controller for step change in

reference speed from 50% of the rated value to the rated value
at 1sec

39
 ©2010 International Journal of Computer Applications (0975 - 8887)
Volume 1 – No. 27

Fig.12 Simulated response of adaptive fuzzy controller for instructions with faster execution speed, programmable code
change in speed for a load change from 0 to 100% of the rated protection, selectable power management modes and low power
value at 1sec. consumption. Port B of the processor receives input from the
position sensors. Based upon the values of hall position sensor
signals, PWM pulses are produced at the Port E Pins 0 to 5.
Performance Comparison
The speed is calculated from the hall sensor inputs using timers.
The performance of the system with fuzzy and adaptive fuzzy
Then the controller routine will decide the new duty cycle. The
controller is measured in terms of rise time and steady state error.
duty cycle of the PWM generated can be varied by changing the
The performance comparison chart for different reference speeds
control word written to the corresponding special function register
for no load condition is given in table 2. Table 3 gives the
in the processor. By varying the duty cycle of the pulses applied
comparison of the performance parameters under loaded
to the inverter, the speed of the PMBLDC motor is controlled
condition.
effectively. In this work, low side pulses are made as PWM pulses
Table 2: Performance comparison under no load whereas high side drivers are kept at active high or low state as
and when required. The flowchart of the control program is given
in figure 13.
Step change Fuzzy Adaptive Fuzzy
in reference controller Controller
speed
Rise Steady Rise Steady
time in state time in state
sec error sec error
in in
% %
0 to 100% of 0.1779 0.866 0.1779
rated value 0.4

0 to 75% of 0.1029 0.355 0.1029 0.177

rated value
0 to 50% of 0.0567 0.333 0.0567
rated value 0.12

0 to 25% of 0.0242 0.053 0.0242 -

rated value 0.053

Table 3: Performance comparison under 100% of the load

torque applied
Step change Fuzzy Adaptive Fuzzy
in reference controller Controller
speed
Rise Steady Rise Steady
time in state error time in state error
sec sec
in % in %
0 to 75% of 0.1700 0.933 0.1700 0.466
rated value
Figure 13 Flowchart of the control algorithm
0 to 50% of 0.0810 0.706 0.0810
rated value 0.333

0 to 25% of 0.0337 0.373 0.0337 -

0.066 CONCLUSION
rated value From the simulation results, it is inferred that wide range of speed
control is possible by means of fuzzy controller. The results are
obtained without peak overshoot and with negligible steady state
error. Moreover it can be noted that the controller design does not
Experimental Implementation requires explicit knowledge of the motor and load characteristics.
The fuzzy controller was implemented practically using
So, it can be concluded that this method of speed control of
dsPIC30F4011 processor and programming it in C language. A
BLDC motor provides accurate results even under ill-defined
six switch inverter was constructed and used to test the controller
nonlinear dynamics. The results of adaptive fuzzy controller are
designed on a 24V BLDC motor. The processor selected has the
better than the fuzzy controller without adaptation. The practical
advantages of Enhanced flash program memory, self-
reprogrammable under software control, limited number of

40
 ©2010 International Journal of Computer Applications (0975 - 8887)
Volume 1 – No. 27

implementations of the designed fuzzy and adaptive fuzzy 5. Gwo-Ruey Yu; Chin-Wen Chuang; Rey-Chue Hwang;
controllers are discussed. “Fuzzy control of brushless DC motors by gray
prediction” IFSA World Congress and 20th NAFIPS
International Conference, Canada, 2007. Joint 9th
REFERENCES Volume , Issue , 25-28 July 2007Page(s):2819 - 2824
1. B.K. Bose, “Power electronics and motor drives-recent vol.5.
technology advances”, Proceedings of the IEEE 6. Hailong Song Yong Yu Ming Yang Dianguo Xu ,
International Symposium on Industrial Electronics (ISIE “A novel SMC-fuzzy speed controller for permanent
2002), Italy, July 8-11,2002, Volume 1, pages 22-25. magnet brushless DC motor” Eighteenth Annual IEEE
2. C.K.Leeand, W.H.Pang, “A Brushless Motor speed Applied Power Electronics Conference and Exposition,
control system using Fuzzy Rules” by Proceedings of 2003. APEC '03, Feb. 2003 Volume: 1, On page(s):
the Fifth IEEE International Conference on Power 281- 285 vol.1
Electronics and Variable, Speed Drives, UK, October 7. N.Senthil Kumar, V.Sadasivam, Muruganandam, M “A
16-28, 1994, Pages 101-106. Low-cost Four-quadrant Chopper-fed Embedded DC
3. Y.H. Bharathi, B.R. Rekha, P. Bhaskar, C.S. Parvathi Drive Using Fuzzy Controller” Electric Power
and A.B. Kulkarni, “Multi-input Fuzzy Logic Controller Components and Systems journal, Volume 35, Issue 8
for Brushless dc Motor Drives”, Defence Science August 2007 , pages 907 - 920
Journal, Vol. 58, No. 1, January 2008, pp. 147-158 8. Dimiter Drainkov, Hans Hellendoorn and Michael
4. Uddin, M.N. Radwan, T.S. Rahman, M.A. , Fuzzy- reinfrank, “An introduction to fuzzy control”, Narosa
logic-controller-based cost-effective four-switch three- publishing house, New Delhi,1996
phase inverter-fed IPM synchronous motor drive
system”, IEEE Transactions on Industry Applications,
Jan.-Feb. 2006, Volume: 42, Issue: 1, page(s): 21- 30.

41