See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/310830301

Speed Control of Separately Excited DC Motor using Fuzzy Neural Model

Reference Controller

Article in International Journal of Instrumentation and Control Systems · October 2012

DOI: 10.5121/ijics.2012.2403

CITATIONS READS
36 4,056

1 author:

Waleed I. Breesam

17 PUBLICATIONS 182 CITATIONS

SEE PROFILE

All content following this page was uploaded by Waleed I. Breesam on 06 September 2018.

The user has requested enhancement of the downloaded file.

 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

SPEED CONTROL OF SEPARATELY EXCITED DC

MOTOR USING FUZZY NEURAL MODEL
REFERENCE CONTROLLER
Waleed I. Hameed1 and Khearia A. Mohamad2
1
Department of Electrical Engineering, University of Basrah, Basrah, IRAQ
[email protected]
2
Department of Electrical Engineering, University of Basrah, Basrah, IRAQ
[email protected]

ABSTRACT
Conventional controllers are generally used to control the speed of the separately excited DC motors in
various industrial applications. It is found to be simple and high effective if the load disturbances is small.
So the drawback of Conventional controllers when high load has been applied to the DC motor. This paper
presents the speed control of a separately excited dc motor using Fuzzy Neural Model Referance controller.
The system has been implemented using Matlab/Simulink software. The simulation results show that
presenting controller give good performance and high robustness in load disturbance.

KEYWORDS
Separately excited DC motor, Chopper circuit, Hysteresis current control, Speed control, Model reference
adaptive controller, Fuzzy neural controller.

1. INTRODUCTION

The speed of DC motors can be adjusted within wide boundaries so that this provides easy
controllability and high performance. DC motors used in many applications such as steel rolling
mills, electric trains, electric vehicles, electric cranes and robotic manipulators require speed
controllers toper form their tasks[1]. The most flexible control is obtained by means of separately
excited DC Motor in which the armature and field circuits are provided with separate sources. For
the armature source a controlled rectifier or chopper is required[2]. Armature voltage control
method is used to vary the speed up to the rated speed and the motor operates in the constant
torque region[3]. Conventional controllers such as PI and PID have been applied to control the
speed of DC motors. The disadvantages of using conventional controllers are that they are
sensitive to variation in the motor parameters and load disturbance. In addition, it is difficult to
tune PI or PID gains to eliminate and reduce the overshoot and load disturbance. In order to avoid
the shortcomings of conventional controllers, researchers applied adaptive control techniques for
DC motor speed control to achieve parameter insensitivity and fast speed response[4]. Model
reference adaptive controller (MRAC), was originally proposed to solve a problem in which the
specifications are given in terms of a reference model that tells how the process output ideally
should respond to the command signal[5]. In particular, the possible use of fuzzy systems in
modeling and control has generated great attention. Unfortunately, it is difficult for systems
designers to obtain optimal fuzzy rules because these are most likely to be influenced by the

DOI : 10.5121/ijics.2012.2403 27
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

intuitiveness of the operators and systems designers. Neural network implementation of fuzzy
systems has been proposed as possible approach for fuzzy systems design. In resulting systems,
which are sometimes called neural-network-based fuzzy systems, fuzzy-neural network or neural-
fuzzy network, will possess the advantage of both types of systems ( Neural network and Fuzzy
system) and overcome the difficulties of each type of system. Fuzzy-Neural is a hybrid system
which functions as fuzzy system with processing mechanism realized by fuzzy-neural network[6].
Recently, the use of fuzzy and neural networks to identify and control nonlinear dynamic systems
has been proposed because they can approximate a wide range of nonlinear functions to any
desired degree of accuracy[7]. The objective of fuzzy neural controller is to develop a back
propagation algorithm such that the output speed of the plant can track the reference command
signal[8]. In this paper, Fuzzy Neural Model Reference Controller has been proposed to control
the Speed of Separately Excited DC Motor. This paper is organized as follows; the model of
separately excited DC motor is given in section 2, modeling of circuit drive based on chopper is
given in section 3, also hysteresis current controller (HCC) is given in section 4, in section 5, the
proposed method has been presented to control the speed of separately excited DC motor.
Simulation results are given in section 6, and finally, conclusions are given in section 7.

2. MODELING OF SEPARATELY EXCITED DC MOTOR

In modeling, the aim is to find the governing differential equations that relate the applied voltage
with the produced torque or speed of the rotor[9]. Figure 1 shows the equivalent circuit with
armature voltage control and the model of a general mechanical system that incorporates the
mechanical parameters of the motor and the mechanism coupled to it[2,10].

Figure 1. Equivalent circuit of a separately exited DC-Motor.

The model of the motor consists of a mechanical equation and electrical equation as determined
in the following equations[1,2]:

where is the applied terminal voltage to the motor, is the emf induced in the armature
winding, is the armature inductance of the motor, is the armature motor current, is the
28
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

armature resistance of the motor, is the motor constant, is the motor speed, is the
internal torque of the motor, is the rotational inertia of motor, is the viscous friction of motor
and the load torque.

Table 1 presents the separately excited DC motor characteristics.

Table 1. DC motor parameters

Rated motor speed 500 R.P.M

Rated terminal voltage 230 V
Armature resistance 0.066 Ω
Armature inductance 0.0065 H
The rotational inertia 2.83 Kg/m²
The viscous friction 0.634 Nm/rad/s
Motor constant 4.212 V/rad
Rated armature current 143 A

Mathematical model described through Eqs.( 1) to (3) is implemented using Matlab/Simulink

software. The simulink model of the separately excited DC motor block is shown in Figure 2.

Figure 2. Simulink model of the separately excited DC motor

3. MODELING OF DRIVE CIRCUIT

The dc converter in the voltage step-down connection is required to convert the fixed voltage into
a variable-voltage / variable-current source for the speed control of the DC motor drive. The
variable DC voltage is controlled by chopping the input voltage by varying the on- and off-times
of a converter, and the type of converter capable of such a function is known as a chopper
[11,12]. A schematic diagram of the chopper is shown in Figure 3. The control voltage to it’s gate
is . The chopper is on for a time , and it’s off time is . It’s frequency of operation is

29
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

and it’s duty cycle is defined as

Figure 3. Chopper schematic.

Assume that the switch is ideal, with zero voltage drop, the average output voltage is given as

where is the source voltage.

Varying the duty cycle changes the output voltage. Note that the output voltage follows the
control voltage, as shown in Figure 4 [12].

Figure 4. Voltage waveforms.

The simulink model of the chopper circuit block is shown in Figure 5.

Figure 5. Simulink model of the chopper circuit

30
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

4. MODELING OF HYSTERESIS CURRENT CONTROLLER

Current control of dc machines is very desirable because by controlling the current, the torque is
controlled. Moreover, current control can be used to prevent large damaging armature currents
during start-up[13]. Hysteresis current controller (HCC) is used to generate switching patterns
required for the chopper circuit by comparing the actual current being drawn by the motor with
the reference current[14]. In an HCC, the armature current is forced to stay within the hysteresis
band determined by the upper and lower hysteresis limits ∆i [3]. Commanded current and actual
current are shown in Figure 6.

Figure 6. Hysteresis current operation.

The realization of HCC is shown in Figure 7. The chopping frequency is a varying quantity,
unlike the constant frequency in the PWM controller[12].

Figure 7. Simulink model of hysteresis current controller

31
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

5. FUZZY NEURAL MODEL REFERENCE CONTROLLER

For the systems with changing parameters, the MRAC method gives a general approach for
adjusting controller parameters so that the closed-loop transfer function will be close to a
prescribed model[2]. Many algorithms employ a reference model to provide closed-loop
performance feedback for synthesizing and tuning a fuzzy controller’s knowledge-base.
Consequently, these algorithms are referred to as a “Fuzzy Model Reference Learning Controller”
(FMRLC). Fuzzy Neural Networks were also used instead of fuzzy systems, these algorithms
denoted as “Fuzzy Neural Model Reference Controller” (FNMRC)[15]. The controller is
designed such that the actual output of the process will track the desired output of the reference
model. The MRAC is extensively used in neurofuzzy controllers due to its simplicity [16]. The
block diagram in Figure 8 show the Speed control of Separately Excited DC Motor Using Fuzzy
Neural Model Reference Controller with online tuned.

Figure 8. Block diagram of MRAC.

5.1. Architecture of FNN

The Architecture of FNN shown in Figure 9[17,18]. FNN considered as a special type of neural
network , this means special connection and node operation. Every layer and every node have its
practical meaning because the FNN has the structure which is based on both the fuzzy rules and
inference[15]. In the following items each layer shown in Figure 9 will be described:

1. Input Layer I: Input layer transmits the input linguistic variables to the output without
changed.
2. Hidden Layer II: Membership layer represents the input values with the following Gaussian
membership functions [15]:

Where and (i=1, 2, ... , n ; j=1, 2,.., m), respectively, are the mean and standard deviation
of the Gaussian function in the term of the input linguistic variable to the node of this
layer.

32
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

3. Hidden Layer III: Rule layer implements the fuzzy inference mechanism, and each node in this
layer multiplies the input signals and outputs the result of the product. The output of this layer is
given as [15]:

Where represent the output of rule layer.

4. Output Layer IV: Layer four is the output layer, and nodes in this layer represent output
linguistic variables. Each node ( o = 1,.., ), which computes the output as:

Where represent the output weight of rule layer.

Figure 9. Architecture of FNN.

5.2. Learning Algorithm for FNN Controller

The parameter of the FNN controller presented in Figure 8 should be adjusted according to the
following equations[19]:

33
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

The goal of the learning algorithm is to minimize the error between the desired output (reference
model) and the motor output. In the present work the Least Mean Square (LMS) function will be
used to define a criterion for the error E.

Where η is the learning rate for each parameter in the system, i=1,2…n and j=1,2…m.

6. SIMULATION RESULTS
The SIMULINK model of Fuzzy Neural Model Reference Controller of separately excited DC
motor is shown in Figure 10. The first input of the FNN is error and the second is change of error.
em is the error between model reference and output of the motor that used as trajectory in
adaptive mechanism. In the primes part, the number of membership functions for each input of
the controller are five. So, the number of weights in the consequence part are also five. The
learning rate for each parameter in the controller are shown in Table 2.

Figure 10. Simulink model of Fuzzy Neural Model Reference Controller of separately excited DC motor.

34
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

Table 2. Learning rate for each parameter in the controller.

0.6 0.4 0.4

With no load conditions, the rotor speed of the motor when Nr=400 r.p.m is shown in Figure 11.
Also the LMS error is shown in Figure 12. The armature current at this speed is shown in Figure
13. A step change in reference speed at t=1.25 seconds from 400 r.p.m. to 200 r.p.m is shown in
Figure 14. Figure 15. show the rotor speed of the motor when a 200 N.m load torque applied at
t=1.5 seconds. The armature current and internal torque at this speed are shown in Figure 16. and
Figure 17. respectively.

450

400

350

300 Reference model

Speed (r.p.m)

250 Motor output

200

150

100

50

0
0 0.5 1 1.5 2
Time (Sec.)

Figure 11. Rotor speed of the motor

35
International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

140

120

100

80
LMS error

60

40

20

0 0.2 0.4 0.6 0.8 1

Time (Sec.)

Figure 12. LMS error

180

160

140

120

100
ia (A)

80

60

40

20

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (Sec.)

Figure 13. Armature current of the motor

36
International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

450

400

350

300
Speed (r.p.m)

250

Reference speed step

200 change from 400 to
200 r.p.m
150

Reference model
100
Motor output
50

0
0 0.5 1 1.5 2 2.5
Time (Sec.)

Figure 14. Rotor speed of the motor with step change in reference speed

450

400

350 Load step change

from 0 to 200 Nm
300 405
Speed (r.p.m)

250

200 400
Reference model
150
Motor output
100 395
1.6 1.8 2 2.2

50

0
0 0.5 1 1.5 2 2.5 3
Time (Sec.)

Figure 15. Rotor speed of the motor with step change in load torque

37
International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

180

160

140

120

100
ia (A)

80

60

40

20

0
0 0.5 1 1.5 2 2.5 3
Time (Sec.)

Figure 16. Armature current of the motor

800

700

600

500
Torque (N.m)

400

300

200

100

0
0 0.5 1 1.5 2 2.5 3
Time (Sec.)

Figure 17. Internal torque of the motor with step change in load torque

38
 International Journal of Instrumentation and Control Systems (IJICS) Vol.2, No.4, October 2012

7. CONCLUSIONS

In this paper, Fuzzy Neural Model Reference Controller of separately excited DC motor has been
presented. Fuzzy neural network has been used to control the speed of the motor. The error
between the model reference output and motor speed output is used as trajectory function to
adapting the primes part and the consequence part of the fuzzy-neural network so that the error
goes toward zero. From simulation results, It is found that the proposed system is robust in that it
eliminates the load disturbances considerably.

REFERENCES

[1] Singari.v.s.rPavankumar, Sande.krishnaveni, Y.B.Venugopal, Y.S.Kishore Babu, “A Neuro- Fuzzy

Based Speed Control of Separately Excited DC Motor”, IEEE Transactions on Computational
Intelligence and Communication Networks, pp. 93-98, 2010.
[2] S. B. Dewan, Gordon R. Slemon, A. Straughen, “Power Semiconductor Drives”, John Wiley & Sons,
1984.
[3] Moleykutty George, Kartik Prasad Basu, Alan Tan Wee Chiat, “Model reference controlled
separately excited DC motor”, Springer, Neural Comput & Application, vol. 19, Issue: 3, PP. 343-
351, 2010.
[4] Adel Elkeshet, “New Developments Using ANFIS for DC Motors Speed Control”, MSc.Thesis,
Electrical and Computer Engineering, Dalhousie University, Canada, April, 2009.
[5] Bogdan M. Wilamowski, J. david Irwin, “Power electronIcs and motor drives” , Taylor and Francis
Group, LLC, USA, 2011.
[6] Jae-Sub Ko, Jung-Sik Choi, Dong-Hwa Chung, “Hybrid Artificial Intelligent Control for Speed
Control of Induction motor”, SICE-ICASE International Joint Conference, Oct. 18-21, in Bexco,
Busan, Korea, PP. 678-683, 2006.
[7] Seong-Hwan Kim, Tae-Sik Park, Ji-Yoon Yoo, Gwi-Tae Park, “Speed-sensorless vector control of an
induction motor using neural network speed estimation”, IEEE Transactions on Industrial Electronics,
Vol.48, Issue: 3, PP.609-614, June 2001.
[8] M. Brown, “Neuro Fuzzy Adaptive Modeling and control”, Prentice Hill, International UK, 1994.
[9] T. Kara, I. Eker, “Nonlinear closed-loop direct identification of a DC motor with load for low speed
two-directional operation”, Springer, Electrical Engineering, vol. 86, Issue: 2, PP. 87-96, 2004.
[10] P.C. Sen, “Thyristor DC Derives”, John Wiley and Sons, NY, 1981.
[11] William Shepherd, Li Zhang, “Power Converter Circuit”, Marcel Dekker, Inc, USA, 2004.
[12] R. Krishnan, “Electric Motor Drives, Modeling, Analysis, and Control”, Prentice Hill, USA, 2001.
[13] Paul C. Krause, Oleg Wasynczuk, Scott D. Sudhoff, “Analysis of Electric Machinery and Drive
Systwms”, IEEE Press, USA, 2002
[14] Basma A. Omar, Amira Y. Haikal, Fayz F. Areed, “An Adaptive Neuro-Fuzzy Speed Controller for a
Separately excited DC Motor”, International Journal of Computer Applications, pp. 29-37, Vol. 39,
No.9, February 2012.
[15] Salam Abdul Hady Abdul Kareem, “Fuzzy Neural And Fuzzy Neural Petri Nets Control For Robot
Arm”, MSc.Thesis, Computers Engineering, Basrah University , September, 2010.
[16] Hasan Nassir Hasan AL-Salih, “Neurofuzzy Controller For The Induction Motor Speed UsingThe
Slip Power Recovery Strategy”, MSc.Thesis, Electrical Engineering, Basrah University, December,
2005.
[17] M. S. Mostafa, M. A. El-Bardini, S. M. Sharaf, M. M. Sharaf, “Fuzzy neural networks for
identification and control of DC drive systems”,IEEE International Conference on Control
Applications, pp. 598-603, Vol.1, 2004.
[18] Ahmed Thamer Radhi, “Power System Protection Using Fuzzy Neural Petri Net”, Ph.D. Thesis,
Electrical Engineering, Basrah University, February, 2012.
[19] Khearia A. Mohamad, “Fuzzy Neural Controller For Multi-Machine Induction Motor Drives”, Ph.D.
Thesis, Electrical Engineering, Basrah University, February, 2009.

39

View publication stats