Optimal Tuning of PI Controller for Speed Control of

DC motor drive using Particle Swarm Optimization

Rohit G. Kanojiya, Student, Y.C.C.E, and P. M. Meshram, Associate professor, Y.C.C.E

Abstract—This paper present a method to determine the optimal in time response, still opens the way of tuning parameters.
tuning of the PI controller parameter on Direct current (DC) Modified Ziegler-Nichols tuning based on Chien-Hrones-
motor drive system using particle swarm optimization (PSO) Reswick (CHR) PID tuning formula [15] for set-point
algorithm, Ziegler-Nichols (ZN) tuning and Modified Ziegler- regulation accommodate the response speed and overshoot.
Nichols (MZN) tuning method. The main objective of this paper
The other method used is a population-based optimization
is to minimize transient response specifications chosen as rise
time, settling time and overshoot, for better speed response of DC method first proposed by Eberhart and Colleagues [1, 2].
motor drive. The speed control of DC motor is done using PI and Particle swarm optimization (PSO) has attractive features like,
PID controllers. Implementation of PID controller for DC motor ease of implementation and the fact that no gradient
speed control is done using ZN and MZN tuning method. For information is required. It can be used to solve a wide array of
PSO algorithm technique, PI controller is used to improve the different optimization problems. Like evolutionary algorithms,
performance of DC motor speed control system. A comparison is PSO technique conducts search using a population of particles,
made on the basis of objective function (rise time, settling time corresponding to individuals. Each particle represents a
and overshoot) from output Step responses. The proposed candidate solution to the problem at hand thus has more
approach had superior features, including easy implementation,
efficiency for finding problem solution.
stable convergence characteristic, and good computational
efficiency. Fast tuning of optimum PI controller parameters Control System Design and Analysis Technologies are widely
yields high-quality solution. Compared with traditional ZN suppress and very useful to be applied in real-time development.
method and MZN method, the proposed method is found indeed Some can be solved by hardware technology and by the advance
more efficient and robust in improving the step response of DC used of software, control system are analyzed easily and detail.
motor drive system. DC Motors can be used in various applications and can be used as
various sizes and rates. Today their uses isn’t limited in the car
Keywords-DC motor, optimal contro, particle swarm applications (electrics vehicle), in applications of weak power
optimization, Ziegler-Nichols tuning method, PI controller, PID using battery system (motor of toy) or for the electric traction
controller. in the multi-machine systems too. The speed of DC motor can
be adjusted to a great extent as to provide controllability easy
I. INTRODUCTION and high performance.
Nowadays several control theories have been developed In this paper, an optimal PI-PSO controller, PID-ZN
significantly; we do see the widely popular use of controller and PID-MZN controller are developed for DC
proportional-integral (PI) and proportional integral-derivative motor speed control. The performance measure to be
(PID) controllers in process control, motor drives, flight minimized contains the following objectives of the PID
control, and instrumentation. The reason of this acceptability controller, that will be studied separately,
is for its simple structure which can be easily understood and 1. Minimize the rise time, time required for system
implemented. Industries too can boast of the extensive use of response to rise from 10% to 90% (over damped); 5%
PI and PID controllers because of its robustness and to 95%; 0% to 100% (Under damped) of the final
simplicity. The past decades witnessed many advancing steady state value of the desired response,
improvements keeping in mind the requirement of the end 2. Minimize the maximum overshoot, Maximum
users. Easy implementation of hardware and software has overshoot is the maximum peak value of the response
helped to gain its popularity. Several approaches have been curve measured from the desired response of the
documented in literatures for determining the PID parameters system, and
of such controllers which is first found by Ziegler- Nichols 3. Minimize the settling time, Time required for
tuning [6]. Genetic Algorithm, neural network, fuzzy based response to reach and stay within 2% of final value.
approach [7, 9], particle swarm optimization techniques [1]-
[5] are just a few among these numerous works. II. DC MOTOR MODEL
In 1942, Ziegler-Nichols presented a tuning formula [6, This DC motor system is a separately excited DC motor
13, 14], based on time response and experiences. Although it [7]-[10], which is often used to the velocity tuning and the
lacks selection of parameters and has an excessive overshoot position adjustment. This paper focuses on the study of DC
motor linear speed control, therefore, the separately excited DC
R. G. Kanojiya and P.M. Meshram are with Department of Electrical
Engineering, Yeshwantrao Chavan College of Engg., Nagpur, India (e-mail:
[email protected]; [email protected]).
978-1-4673-2043-6/12/$31.00 ©2012 IEEE

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.
motor is adopted. Make use of the armature voltage control Tm (s) = Bω(s) + JSω(s) = K T Ia (s) (6)
method to control the DC motor velocity, the armature voltage Figure 2 describes the DC motor armature control system
controls the distinguishing feature of method as the flux fixed, function block diagram from equations (1) to (6).
is also a field current fixedly. The control equivalent circuit of
the DC motor by the armature voltage control method is shown
in Fig. 1.

Figure 2. DC Motor Armature Voltage Control System function block

diagram.

The transfer function of DC motor speed with respect to the

input voltage can be written as follows,

ω(s) KT
G(s) = = (7)
E a (s) (La (s) + R a )(JS + B) + K b K T
Figure 1. The control Equivalent Circuit of DC Motor using the Armature
Voltage Control. III. PI AND PID CONTROLLER
Where,
A. PI Controller
R a : Armature resistance; Proportional-Integral (PI) controllers [6, 14] are one of the
La : Armature inductance; most applicable controllers in different industries. The main
important need in application of these controllers is their
ia : Armature current; parameters tuning in order to gain desired result. So an
if : Field current; accessible method with high accuracy and speed has to be used
ea : Input voltage; for determination of these control parameters (Kp, Ki). The
control architecture used for PI controller is shown by figure 3.
e b : Back electromotive force (EMF);
Tm : Motor torque;
ω: An angular velocity of rotor;
J: Rotating inertial measurement of motor bearing;
K b : EMF constant;
K T : Torque constant;
B : Friction constant.
Because the back EMF e b is proportional to speed ω directly, Figure 3. Block diagram of PI controller for PI-PSO controller.
then
PI-controllers have been applied to control almost any process
dθ one could think of, from aerospace to motion control, from
e b (t) = K b = K b ω(t) (1)
dt slow to fast systems. With changes in system dynamics and
Making use of the KCL voltage law can get variation in operating points PI-controllers should be retuned
on a regular basis.
di a (t)
ea (t) = R a ia (t) + La + eb (t) (2) ⎡ t ⎤
dt 1
From Newton law, the motor torque can be obtained as
U(t) = K P ⎢e(t) +
⎢ Ti ∫
e(τ)d(τ) ⎥
⎥
(8)
⎣ 0 ⎦
Where u(t) and e(t) denote the control and the error signals
d 2 θ(t) dθ
Tm (t) = J +B = K T ia (t) (3) respectively, and K P and Ti are the parameters to be tuned.
dt dt
Take (1), (2), and (3) into Laplace transform, respectively, the The corresponding transfer function is given as,
equations can be formulated as ⎡ 1 ⎤
K(s) = K P ⎢1 + ⎥ (9)
⎣ Ti (s) ⎦
E a (s) = (R a + La S)Ia (s) + E b (s) (4)

E b (s) = K b ω(s) (5)

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.
B. PID Controller can be performed to extract the parameters for the approximate
Proportional-Integral-Derivative (PID) controllers [6]–[11] model (12).
are widely used in industrial control systems because of the
reduced number of parameters to be tuned. They provide
control signals that are proportional to the error between the
reference signal and the actual output (proportional action), to
the integral of the error (integral action), and to the derivative
of the error (derivative action), namely
⎡ 1
t
d ⎤
U(t) = K P ⎢e(t) +
⎢
⎣
Ti
0
∫
e(τ)dτ + Td
dt
e(t) ⎥
⎥
⎦
(10)

Figure 5. Response Curve for Ziegler-Nichols Method.

Td is the parameters to be tuned. The corresponding transfer
function is given as For instance, if the step response of the plant model can be
measured through an experiment, the output signal can be
recorded as sketched in Fig. 5, from which the parameters of
⎡ 1 ⎤
K(s) = K p ⎢1 + + Td (s) ⎥ (11) k, L, and T (or a, where a = kL/T ) can be extracted by the
⎣ Ti (s) ⎦ simple approach shown. More sophisticated curve fitting
These functions have been enough to the most control approaches can also be used. With L and a, the Ziegler–
processes. Because the structure of PID controller is simple, it Nichols formula in Table 1 can be used to get the controller
is the most extensive control method to be used in industry so parameters.
far. The PID controller is mainly to adjust an appropriate
proportional gain ( K P ), integral gain ( K I ), and differential
TABLE I. ZIEGLER-NICHOLS TUNING FIRST METHOD
gain ( K D ) to achieve the optimal control performance. The
Controller type Kp Ti Td
PID controller system block diagram of this paper is shown in P T/L
Figure 4. PI 0.9T/L L/0.3
PID 1.2T/L 2L 0.5L

B. Modified Ziegler-Nichols Tuning method

Modified Ziegler-Nichols tuning using Chien–Hrones–
Reswick (CHR) tuning algorithm emphasizes on set-point
regulation. . In addition one qualitative specification on the
response speed and overshoot can be accommodated.
Compared with the traditional Ziegler–Nichols tuning
formula, the CHR method uses the time constant T of the plant
explicitly. The CHR PID controller tuning formulas are
Figure 4. Block diagram of PID controller PID-ZN and PID-MZN summarized in Table 2 for set-point regulation.
controller.
TABLE II. MODIFIED ZIEGLER-NICHOLS TUNING SECOND
IV. ZIEGLER-NICHOLS METHOD METHOD

Controller type Kp Ti Td
A. Empirical method P 0.7/a
This method is applied to plants with step responses of the PI 0.6/a T
form displayed in Fig. 5. This type of response is typical of a PID 0.95/a 1.4T 0.47T
first order system with transportation delay. The response is
characterized by two parameters, L the delay time and T the Here the parameters k, L, and T are obtained from the
time constant. These are found by drawing a tangent to the response curve of Fig. 5. With K I = K P / Ti and K D = K P * Td .
step response at its point of inflection and noting its
intersections with the time axis and the steady state value. The
plant model is therefore,
Ke−sL
G(s) = (12)
TS + 1
Ziegler and Nichols derived the following control parameters
based on this model. In real-time process control systems, a
large variety of plants can be approximately modeled by (12).
If the system model cannot be physically derived, experiments Figure 6. Block diagram of DC motor control system used by PID-ZN and
PID-MZN controller.

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.
 Thus the values of K P , Ki and Kd are obtain from tables 1
and 2 to form the transfer function for PID controller. The PID-
ZN and PID-MZN controller are made by using table 1, table 2
and applied for speed control of DC shown by block diagram
(Fig. 6)

V. PARTICLE SWARM OPTIMIZATION

According to the background of PSO and simulation of
swarm of bird, Kennedy and Eberhart developed a PSO
concept [2]-[5]. Namely, PSO is basically developed through
simulation of bird flocking in two-dimension space. The
position of each agent is represented by XY axis position and
also the velocity is expressed by vx (the velocity of X axis) and Figure 7. Concept of Modification of searching point by PSO.
vy (the velocity of Y axis). Modification of the agent position Here,
is realized by the position and velocity information.
Bird flocking optimizes a certain objective function. Each Sk : Current searching point,
agent knows its best value so far (pbest) and its XY position. Sik +1 : modified searching point,
This information is analogy of personal experiences of each
agent. Moreover, each agent knows the best value so far in the V k : Current velocity,
group (gbest) among pbests. This information is analogy of Vik +1 : modified velocity,
knowledge of how the other agents around them have
Vpbest : Velocity based on pbest,
performed. Namely, Each agent tries to modify its position
using the following information: Vgbest : Velocity based on gbest.
- the current positions (x, y),
- the current velocities (vx, vy),
- the distance between the current position and pbest
- the distance between the current position and gbest
This modification can be represented by the concept of
velocity. Velocity of each agent can be modified by the
following equation:
Vik +1 = WVik + C1 rand1 × (pbest i − Sik ) + …
Figure 8. Searching concepts with Agents in solution space by PSO.
C2 rand 2 × (gbest − Sik ) (13)
VI. RESULT AND ANALYSIS
Where,
A standard test model as considered is taken for study of
Vik : Velocity of agent i at iteration k, DC motor with Z-N tuning controller. The test model below
W: weighting function, shown is completely designed in SISO tool. Fig. 6 shows the
C j : weighting factor, block diagram of DC motor driving an inertial load.
rand: random number between 0 and 1,
TABLE III. DC MOTOR PARAMETERS
Sik : Current position of agent i at iteration k,
pbest i : pbest of agent i, Parameters Motor 1 Motor 2
gbest: gbest of the group. Armature Resistance Ra 2 2
(Ω )
Using the above equation, a certain velocity, which
gradually gets close to pbest and gbest can be calculated. The Armature Inductance La 0.5 0.5
(H)
current position (searching point in the solution space) can be Moment of Inertia J 0.02 1.2
modified by the following equation: (Kgm2)
Friction constant B 0.2 0.2
Sik +1 = Sik + Vik +1 (14) (Nms)
Torque constant KT 0.015 0.2
Fig.7 shows a concept of modification of a searching point (Nm/A)
by PSO and Fig. 8 shows a searching concept with agents in a EMF constant KB 0.01 0.2
solution space. Each agent changes its current position using (Vs/rad)
the integration of vectors as shown in Fig. 7.
From the state equation (refer (1), (2), (3)) previously, we
can construct the model with the environment MATLAB

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.
(R2010a) Simulink. The model of the DC motor in Simulink is Hrones–Reswick (CHR) tuning algorithm. The step responses
shown in Fig. 10. The various parameters of the DC motor are are shown by Fig. 12 and 13.
shown in Table 3.
C. Particle Swarm Optimization
A. Empirical method Flow Chart as shown in Fig. 11 is used for MATLAB
Flow Chart as shown in Fig. 9 is used for MATLAB coding to find the PI controller parameters and function calling
coding to find the PID controller parameters and to get DC of DC motor transfer function is done to find the close loop
unit step response.
motor Close loop unit step response of the overall transfer
function.

Figure 11. The Flow Chart of Particle swarm optimization Algorithm.

Figure 9. The Flow Chart of Ziegler-Nichols Tuning Method.
The Step responses by MATLAB coding, using the PSO
The controller is connected in cascaded fashion and step flow chart are shown by 12 and 13.
responses for different motors are shown by Fig. 12 and 13. Close loop Step Response
1.4
B. Modified Ziegler-Nichols Tuning Method ZN
1.2 Modified ZN
PI-PSO
1
Amplitude

0.8

0.6

0.4

0.2

0
0 2 4 6 8 10 12
Figure 10. Matlab simulink model for Armature control of DC motor. Time (sec)
Using MATLAB R2010a we can design compensator by Figure 12. Motor 1.Close loop step response with PID-ZN controller, PID-
Ziegler-Nichols open loop tuning method based on Chien– MZN controller and PI-PSO controller.

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.
 Close loop Step Response REFERENCES
1.6
ZN
[1] Y. Valle, G. Venayagamoorthy, S. Mohagheghi, J. Hernandez and R.
1.4 Harley, "Particle swarm optimization: Basic concepts, variants and
Modified ZN
applications in power systems”, IEEE trans. on evolutionary
PI-PSO computation, vol. 12, no. 2, April 2008.pp. 171-195.
1.2
[2] J. Kennedy, “The particle swarm: Social adaptation of knowledge,” in
1 Proc. IEEE Int. Conf. Evol. Comput., Apr. 1997, pp. 303–308.
Amplitude

0.8 [3] H. Zhu, Yi. wang and S. Lee,” Particle Swarm Optimization (PSO) for
the constrained portfolio optimization problem,” Expert Systems
0.6 with Applications 38 (2011), ELSVIER LTD., 2011, pp. 10161–10169.
[4] Y. Dong, J. Tang, B. Xu and D. Wang,” An Application of swarm
0.4 optimization to Non-Linear programming,” Computers and Mathematics
with applications 49 (2005), ELSVIER LTD.,2005, pp. 1655-1668.
0.2
[5] S. Wahsh and A. Elwer,” Improved performance of Permanent Magnet
0 synchronous motor by using Particle swarm optimization techniques,” in
0 2 4 6 8 10 12 Proc.of 2007 IEEE International Conference on Robotics, 2008, pp.
Time (sec) 2095-2100.
[6] J. C. Basilio and S. R. Matos, ”Design of PI and PID Controllers With
Figure 13. Motor 2. Close loop step response with PID-ZN controller, PID-
MZN controller and PI-PSO controller. Transient Performance Specification”, ,” IEEE Trans. Education, vol.
45, Issue No. 4, 2002, pp. 364-370.
[7] O. Montiel, R. Sepulveda, P. Melin, O. Castillo, M. A. Porta and I. M.
TABLE I. TRANSIENT RESPONSE OF MOTOR 1 Meza, “Performance of a Simple Tuned Fuzzy Controller and a PID
Controller on a DC Motor,” in Proc. IEEE Symposium on Foundation of
Method Rise time Maximum Settling Computational Intelligence (FOCI-2007), 2007, pp. 531-537.
(sec) overshoot time [8] N. Kamaruddin, Z. Janin, Z. Yusuf and M. N. Taib, ”PID Controller
(%) (sec) Tuning for Glycerin Bleaching Process Using Well-Known Tuning
PID-ZN 0.312 27.9 2.27 Formulas- A Simulation Study,” in Proc. of 35th Annual Conference of
controller IEEE on Industrial Electronics (IECON-2009), 2009, pp. 1682-1686.
PID-MZN 0.074 14.5 0.439 [9] Y. Ma, Y. Liu and C. Wang Elissa, “Design of Parameters Self-tuning
controller Fuzzy PID Control for DC Motor,” in Proc. of Second International
PI-PSO 0.3907 0.042 0.6467 Conference on Industrial Mechatronics and Automation (ICIMA), vol. 2,
controller 2010, pp. 345-348.
[10] K. A. Naik and P. Shrikant, “Stability Enhancement of DC Motor using
TABLE II. TRANSIENT RESPONSE OF MOTOR 2 IMC Tuned PID Controller,” (IJAEST) International Journals of
Advanced Engg. Science and Technologies, vol. 4, Issue No. 1, 2011,
Method Rise time Maximum Settling pp. 092-096.
(sec) overshoot time [11] W. P. Aung, “Analysis on Modeling and Simulink of DC Motor and its
(%) (sec) Driving System Used for Wheeled Mobile Robot,” World Academy of
PID-ZN 0.377 51.8 5.68 Science, Engineering and Technology 32, 2007, pp. 299-306.
controller [12] G. Haung and S. Lee,” PC Based PID speed control of DC motor,” in
PID-MZN 0.359 7.7 1.07 Proc. of International Conference on Audio, Language and Image
controller Processing (ICALIP-2008), 2008, pp. 400-407.
PI-PSO 0.1678 0 0.2652 [13] J. G. Ziegler and N. B. Nichols, ”Optimum setting for automatic
controller controllers,” Trans. ASME, vol. 64, 1942, pp. 759–768.
[14] K. Astrom and T. Hagglund, PID Controller: Theory, Design and
Tuning, 2nd ed, Library of Congress Cataloging-in-Publication Data,
1994, pp. 120–134.
VII. CONCLUSION [15] D. Xue, Y. Chen and D. P. Atherton, Linear Feedback control, Society
In this paper, the tuning parameters of PID controller is of Industrial and Applied Mathematics, 2007, ch. 6.
designed using traditional Ziegler-Nichols modified Ziegler-
Nichols tuning algorithms and PI controller is designed using
particle swarm optimization The results of all the three
methods are checked by MATLAB coding as well as
simulation. The speed of a three different DC Motor
parameters is controlled by means of these three controllers.
According to the results of the computer simulation, the
particle swarm optimization tuned controller efficiently is
better than the Ziegler-Nichols and Modified Ziegler-Nichols
controller. The PI-PSO controller is the best controller which
presented satisfactory performances for the objectives and it is
the most robust technique.

Authorized licensed use limited to: Universiti Malaysia Perlis. Downloaded on January 21,2022 at 09:15:58 UTC from IEEE Xplore. Restrictions apply.