ANN - Control System DC Motor

Phan Quoc Dzung Le Minh Phuong

Faculty of Electrical & Electronic Engineering Faculty of Electrical & Electronic Engineering
HCMC University of Technology HCMC University of Technology
Ho Chi Minh City - Vietnam Ho Chi Minh City - Vietnam
[email protected] [email protected]

Abstract—This paper introduces the new ability of Artificial control complicated electro- mechanic system such as DC
Neural Networks (ANNs) in estimating speed and controlling motor and synchronous machines [5].
the separately excited DC motor. The neural control scheme
consists of two parts. One is the neural estimator which is used to To train ANNs, we have to determine the input and output
estimate the motor speed. The other is the neural controller datasheets first, and then design the ANNs net by optimizing
which is used to generate a control signal for a converter. These the number of hidden layers, the number of neural of each layer
two neurals are training by Levenberg-Marquardt back- as well as the input/output number and the transfer function.
propagation algorithm. ANNs are the standard three layers feed- The following is to find the ANNs net learning algorithm.
forward neural network with sigmoid activation functions in the
input and hidden layers and purelin in the output layer. ANNs are trained relying on two basic principal: supervisor
Simulation result are presented to demonstrate the effectiveness and unsupervisor. According to supervisor, ANNs learn the
of this neural and advantage of the control system DC motor with input/ output data (targets) before being used in the control
ANNs in comparison with the conventional scheme without system.
ANNs. In this paper, the author would like to present the new
ANNs application in speed estimating and controlling
Keywords:DC motor, artifical neural networks, control system.
separately excited DC motor. The motor speed is controlled by
forecasting method and forecasting task which ANNs
I. INTRODUCTION undertake from the terminal voltage parameter, armature
Nowadays, the field of electrical power system control in current and a reference speed.
general, and motor control in particular have been researching
broadly. The new technologies are applied to these in order to II. DC MOTOR CONTROL MODEL WITH ANNS
design the complicated technology system. One of these new
The DC motor is the obvious proving ground for advanced
technologies is Artificial Neural Networks (ANNs) which base
on the operating principle of human being nerve neural. control algorithms in electric drives due to the stable and
straight forward characteristics associated with it. It is also
There are a number of articles that use ANNs applications to ideally suited for trajectory control applications as shown in
identify the mathematical DC motor model. And then this reference [1-3]. From a control systems point of view, the DC
model is applied to control the motor speed [1]. They also uses motor can be considered as SISO plant, thereby eliminating the
inverting forward ANN with two input parameters for adaptive complications associated with a multi-input drive system.
control of DC motor [4].
However, these researches were not interested in the ability A. Matematical model of DC motor
of forecasting and estimating the DC motor speed. The separately excited DC motor is described by the
following equations:
ANNs are applied broadly because of the following special
qualities: KFωp(t) = – Raia(t) – La[dia(t)/dt] + Vt(t) (1)
1. All the ANN signals are transmitted in one KFia(t) = J[dωp(t)/dt] + Bωp(t) + TL(t) (2)
direction, the same as in automatically control
system. where,
ωp(t) - rotor speed (rad/s)
2. The ability of ANNs to learn the sample.
Vt(t) - terminal voltage (V)
3. The ability to creating the parallel signals in
Analog as well as in the discrete system. ia(t) - armature current (A)
4. The adaptive ability. TL(t) - load torque (Nm)
With the special qualities mentioned above, ANNs can be J - rotor inertia (Nm2)
trained to display the nonlinear relationships that the KF - torque & back emf constant (NmA-1 )
conventional tools could not implemented. It also is applied to

(Provide 1cm space here for IEEE catalogue No.)

 B - viscous friction coefficient (Nms)
Ra - armature resistance (Ω)
La - armature inductance (H)
From these equations we can create mathematical model of
the DC motor. The model is presented in Figure1.
Where,
Ta -Time constant of motor armature circuit and Ta=La/Ra (s)
Figure 3. DC motor control model with ANNs
Tm – Mechanical time constant of the motor Tm=J/B (s)
D. The structure and the process of learning ANNs.
ANNs have been found to be effective systems for learning
discriminates for patterns from a body of examples [5].
Activation signals of nodes in one layer are transmitted to the
next layer are through links which either attenuate or amplify
the signal.
An ANNs are trained to emulate a function by presenting it
with a representative set of input/output functional patterns.
The back-propagation training technique adjusts the weights in
all connecting links and thresholds in the nodes so that the
Figure 1. The mathematical model of a separately DC motor
difference between the actual output and target output are
minimized for all given training patterns [1].
B. The conventional control systems of DC motor
In designing and training an ANN to emulate a function, the
There are different methods to synthesize a control systems only fixed parameters are the number of inputs and outputs to
of DC motor, but for a comparison with method used ANNs the ANN, which are based on the input/output variables of the
authors presented a conventional control system of DC motor, function. It is also widely accepted that maximum of two
where the regulator current and regulator speed are synthesized hidden layers are sufficient to learn any arbitrary nonlinearity
by Bietrage-optimum to reduce the over-regulation [6]. [4]. However, the number of hidden neurons and the values of
learning parameters, which are equally critical for satisfactory
learning, are not supported by such well established selection
criteria. The choice is usually based on experience. The
ultimate objective is to find a combination of parameters which
gives a total error of required tolerance a reasonable number of
training sweeps [1,2,3].

Figure 2. Conventional model of control system DC motors

In the conventional model current and voltage sensors are

very important elements and they are a main role during
regulation of speed alongside with regulator current and
regulator speed.

C. The control system of DC motor using ANNs

The control system of DC motor using ANNs is presented in
the Figure3, where ANN1, ANN2 are trained to emulate a
function: ANN 1 to estimate the speed, ANN2 to control
terminal voltage.

Figure 4. Structure of ANN1

The ANN1 and ANN2 structure is shown in Figure4, and

Figure5. It consists of an input layer, output layer and one
hidden layer. The input and hidden layers are tansig-sigmoid
1.5
activation functions, while the output layer is a linear function.
Three inputs of ANN1 are a reference speed ωr(k), a terminal 1
voltage Vt(k-1) and an armature current ia(k-1). And output of
ANN1 is an estimated speed ωp*(k).
0.5

The ANN2 has four inputs: reference speed ωr(k), a terminal

speed
voltage Vt(k-1), an armature current ia(k-1) and an estimated 0
speed ωp* (k) from ANN1. The output of ANN2 is the control
signal for converter Alpha. -0.5

-1

-1.5
0 2 4 6 8 10 12 14 16
time(s)

Figure 7. The target speed DC motor

For ANN2 Alpha = f[ωr(k),Vt(k-1),ia(k-1),ωp*(k)] (4)

The training program of ANN1 and ANN2 are written in the
Neural Network of Matlab program under m-file and it uses the
Levenberg – Marquardt back propagation. There is no any
reference that mention to the optimal number of neural in each
layer, so collecting the neural networks becomes more
complicated.
In order to choose the optimal number of neural, the neural
Figure 5. Structure of ANN2
network is trained by m-file program, reducing the number of
The ANNs are trained off-line using inputs patterns of neural in ANNs hidden layer until the learning error can be
accepted.
ωr(k),Vt(k),ia(k) - for ANN1, and of ωr(k),Vt(k),ia(k),ωp*(k) –
for ANN2. The reference speed – training data ωr(k) is shown The ANNs and the training effort are briefly described by the
in Figure 6. Other inputs patterns are taken from actual following statistics.
hardware and the corresponding output target is shown in
Figure 7. TABLE 1 THE RESULTS OF THE ANN TRAINING
The motor speed is estimated by the trained ANN predictor
Network ANN1 ANN2
as :
Number of input 3 4
For ANN1 ωp*(k) = f[ωr(k),Vt(k-1),ia(k-1)] (3)
Number of output 1 1
1.5
Number of hidden layer 1 1

1 Number of hidden neurons 3 4

Number of training patterns 1215 1215

0.5
Number of training sweeps 5000 5000
speed

0 Learning error 1e-7 1e-8

-0.5

III. SIMULATION OF CONTROL SYSTEMS DC MOTOR

-1
To simulation the conventional control system and control
system with ANNs we used A Simulink/Matlab program the
-1.5
0 2 4 6 8 10 12 14 16 toolbox of Neural-network. The DC motor , which is used in
time(s) models has the follows parameter 5HP, 240V, 1750 RPM,
field 150V, J=0.02215 Nm2, KF=1.976 NmA-1, B=0.002953
Figure 6. The Reference speed DC motor Nms, Ra=11,2 Ω, La=0.1215 H.
 To compare the quality of two control systems we consider
different operating modes of the DC motor:
a) When parameters of the DC motor are constant.
The start of DC motor and regulation of DC motor speed for
two models of control systems is shown in Figure 7,8,9,10. The
results show that both models have a good performance and the
control quality is same for two models.
2) When parameters of the DC motor are not constant.
Figures 11,12,13 show a start of the DC motor for
conventional model and model with ANNs when mechanical
time constant of the motor reduces and gives : Tm1=75%Tm;
Tm2=50%Tm ; Tm3=30%Tm . Where Tm is the mechanical time
constant of the motor when parameters of the DC motor are
constant.
Figures 14,16 show a regulation of speed for conventional
model when Tm2=50%Tm ; Tm3=30%Tm . Figures 15,17 show a
regulation of speed for model with ANNs when Tm2=50%Tm ;
Tm3=30%Tm . Obviously, that model with ANNs has the better Figure 10. Reverse the DC – motor with ANNs
performance when Tm is variation. Thus, in the conventional
model a speed of the DC motor is fluctuating and the system
may be instability.

IV. THE SIMULATING RESULTS

Figure 11. Regulation of DC motor speed - Conventional model

Figure 8. Starting the DC motor – Conventional model

Figure 9. Starting the DC motor – Model with ANNs

Figure 12. Regulation of DC motor speed – Model with ANNs

Figure 13. Starting the DC motor. Tm1 =75%Tm Figure 16. Regulation of DC motor speed - Conventional model. Tm2=50%Tm

Figure 14. Starting the DC motor. Tm2 =50%Tm Figure 17. Regulation of DC motor speed – Model with ANNs. Tm2 =50%T m

Figure 15. Starting the DC motor.T m3=30%T m Figure 18. Regulation of DC motor speed - Conventional model.Tm3=30%Tm
 Figure 19. Regulation of DC motor speed – Model with ANNs.Tm3 =30%Tm.

V. CONCLUSION
The DC motor has been successfully controlled using an
ANN. Two ANNs are trained to emulate functions: estimating
the speed of DC motor and controlling the DC motor,
Therefore , and so ANN can replace sensors speed in the model
of the control systems. Using ANN, we don’t have to calculate
the parameters of the motor when designing the system control.
It is shown an appreciable advantage of control system using
ANNs above the conventional one, when parameters of the
DC motor is variable during the operation of the motors. The
satisfied ability of the system control with ANNs is much
better than the conventional system control. ANN application
can be used in adaptive controlling in the control system
machine with complicated load. Nowadays, in order to
implement the control systems using ANNs for DC motor on
actual hardware, the ANN micro processor is being used.

REFERENCE
[1] S.Weerasoorya and M.A Al-Sharkawi “Identification and control of a
DC Motor using back-propagation neural networks” IEEE transactions on
Energy Conversion, Vol.6, No.4 pp663-669,December 1991
[2] D.Psaltis A. Sideris and A.A. Yamamura, “A Multilayered Neural
Network Controller”, IEEE Control System Magazine, pp.17-20,1988.
[3] B. Kosko, Neural Networks and Fuzzy Systems, Prentice-Hall
International Inc,1992
[4] E. Levin, R. Gewirtzman, and G.F. Inbar, “Neural Network
Architecture for Adaptive System Modellinggand Control”, Neural Networks,
No 4(2) pp 185-191,1991.
[5] G.M. Scott, Knowledge- Based Artificial Neural Networks for Process
Modelling and Control, PhD. Thesis, University of Wisconsin,1993.
[6] S.G German-Gankin – The computing modeling for power –
electronic systems in Matlab, 2001