Journal of Electrical Engineering & Technology (2021) 16:955–961

https://doi.org/10.1007/s42835-021-00660-5

ORIGINAL ARTICLE

Grey Wolf Optimizer Based PID Controller for Speed Control of BLDC

Motor
Pallav Dutta1   · Santanu Kumar Nayak2

Received: 14 July 2020 / Revised: 28 December 2020 / Accepted: 7 January 2021 / Published online: 19 January 2021
© The Korean Institute of Electrical Engineers 2021

Abstract
A BLDC motor is superior to a brushed DC motor, as it replaces the mechanical commutation unit with an electronic one;
hence improving the dynamic characteristics, efficiency and reducing the noise level marginally. Maximum BLDC motor
drives use PID controller to control the speed of the machine; because it is simple in structure, relatively cheaper and exhibits
good performance. But the main problem associated with PID controller is adjusting its parameters during implementation.
In recent works, it has been observed that Particle Swarm Optimization (PSO) technique showed good performance in tuning
PID controller. For this purpose, in this article, “Grey Wolf Optimization” (GWO) algorithm is introduced; which is used
to optimally tune the PID controller parameters. The objective of this article is to compare the results obtained for tuning of
PID controller based on of GWO and PSO technique and a conclusion has been derived that the proposed approach yields
superior dynamic performance for BLDC motor.

Keywords  Brushless DC motor · Grey wolf optimization · PID controller · Particle swarm optimization · Soft computing
technique

1 Introduction of their advantages [1, 2] over the other type of motors for
these sophisticated applications. As a BLDC motor, does not
Modern speed control techniques for variable speed drives require a commutator-brush segment so it is compact, more
have been changed drastically as compared to their conven- efficient, and generate very less noise when in operation [3,
tional counterparts. Before evolution of power electronics, 4]. It exhibits excellent dynamic characteristics on load vari-
conventional (e.g. field and armature flux) control methods ation. One can note that in industrial applications more than
were being used in DC motor drives. Then power electron- 90% of variable speed drives use PID controllers only. Even
ics-based drives gained popularity. For industrial drive some advanced hybrid control techniques such as Fuzzy-
applications, closed loop control techniques were intro- Neural Networks, Fuzzy-Ants Colony, Fuzzy-Genetic Algo-
duced and PI, PID controllers were used along with power rithm, Fuzzy-Swarm, etc give better performance [5] still
electronic converters [1]. Now-a-days for more sophisticated PID is preferred because it is cheap and has simple structure
applications such as space craft and aeronautical engineer- [6, 7]. For a particular application, the performance of a PID
ing, biomedical instrumentation, robotic application etc.; controller depends upon its parameters ­(KP, ­KI, ­KD). Gener-
the performance of these conventional DC motors is not up ally, the values of these parameters are evaluated by tuning
to the mark. In this work BLDC motor is selected because methods like Ziegler-Nichol optimization method or Cohen-
Coon method, but these methods have some restrictions [8].
Advanced optimization algorithms like PSO, Genetic Algo-
* Pallav Dutta rithm etc. are more efficient and exhibit better steady state
pallav.dutta@aol.com characteristics [9]. In this article, Grey Wolf Algorithm is
Santanu Kumar Nayak applied to the PID controller, and the results are compared
santanu.nayak.1991@gmail.com with PSO algorithm; which is a popular optimization algo-
1 rithm [10]. In the later sections of this article mathemati-
Department of Electrical Engineering, Aliah University,
Kolkata 700160, India cal modeling of BLDC motor and PID controller have been
2 discussed. Third and fourth sections are describing PSO and
Technology Analyst, Infosys Ltd, Hyderabad 500088, India

13
Vol.:(0123456789)

956 Journal of Electrical Engineering & Technology (2021) 16:955–961

GWO algorithms respectively. The subsequent section deals So,

with the selection of fitness functions. In the last section
authors, have given the comparative study in case of tran-
Mib + Mic = −Mia (5)
sient and steady state response in speed variations through Substituting (5) in (4) the state space form of BLDC
simulated results. motor is obtained as-

⎡ Vas ⎤ ⎡ Rs 0 0 ⎤⎡ ia ⎤ ⎡L−M 0 0 ⎤⎡ ia ⎤ ⎡ ea ⎤
⎢ Vbs ⎥ = ⎢ 0 Rs 0 ⎥⎢ ib ⎥ + d ⎢ 0 L − M 0 ⎥⎢ ib ⎥ + ⎢ eb ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ dt ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎣ Vcs ⎦ ⎣ 0 0 Rs ⎦⎣ ic ⎦ ⎣ 0 0 L − M ⎦⎣ ic ⎦ ⎣ ec ⎦
2 Mathematical Modelling
(6)
2.1 BLDC Motor Modelling
2.2 PID Controller Design
Assuming the resistances of all the phase windings of a
BLDC motor are equal, the phase voltages can be repre- A PID controller having a parallel structure is shown in
sented by equation-1 [11]. Fig. 1.

⎡ Vas ⎤ ⎡ Rs 0 0 ⎤⎡ ia ⎤ ⎡ Laa Lab Lac ⎤⎡ ia ⎤ ⎡ ea ⎤ KI

PID = Kp + + Kd S (7)
⎢ Vbs ⎥ = ⎢ 0 Rs 0 ⎥⎢ ib ⎥ + d ⎢ Lba Lbb Lbc ⎥⎢ ib ⎥ + ⎢ eb ⎥ S
⎢ ⎥ ⎢ ⎥⎢ ⎥ dt ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎣ Vcs ⎦ ⎣ 0 0 Rs ⎦⎣ ic ⎦ ⎣ Lca Lcb Lcc ⎦⎣ ic ⎦ ⎣ ec ⎦
(1) The controller processes the error signal directly.
Assume that all the self-inductance of each phase wind- The error signal is obtained from the difference of ref-
ings are equal. Similarly, all the mutual inductances are erence speed and actual speed. When in design stage,
equal. the controller parameters ­K p , ­K i and ­K d depend upon
the closed loop feedback system which are to be chosen
Laa = Lbb = Lcc = L (2) from a long range of values. Practically realizing an ideal
PID controller is not possible. An LPF may be used to
Lab = Lba = Lcb = Lbc = Lca = Lac = M (3) reduce noise to some extent [12]. So, by using advanced
optimization algorithms, response closer to ideal can be
Substituting (2) and (3) in (1):
achieved (Fig. 2).
⎡ Vas ⎤ ⎡ Rs 0 0 ⎤⎡ ia ⎤ ⎡ M L L ⎤⎡ ia ⎤ ⎡ ea ⎤
⎢ Vbs ⎥ = ⎢ 0 Rs 0 ⎥⎢ ib ⎥ + d ⎢ L M L ⎥⎢ ib ⎥ + ⎢ eb ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ dt ⎢ ⎥⎢ ⎥ ⎢ ⎥ 3 Particle Swarm Optimization
⎣ Vcs ⎦ ⎣ 0 0 Rs ⎦⎣ ic ⎦ ⎣ L L M ⎦⎣ ic ⎦ ⎣ ec ⎦
(4)
Inspired by some natural phenomena and animal behavior,
For a balanced three phase stator winding, at any instant
many meta-heuristic optimization concepts like Genetic
summation of all the phase currents is zero [13].
Algorithm, PSO, Ant colony algorithm etc. have been devel-
ia + ib + ic = 0 oped and proved their efficiency in solving complex opti-
mization problems. Among those PSO is one of the most

Fig. 1  Closed loop control

of a BLDC motor using PID
controller

13
Journal of Electrical Engineering & Technology (2021) 16:955–961 957

4 Grey Wolf Optimization

PWM BLDC
DC
inverter motor
source
Grey wolf optimization is a new meta-heuristic optimiza-
tion technique which is purposed by Seyedali Mirjalili,
PWM Seyed Mohammad Mirjalili, and Andrew Lewis (2014)
modulator [21]. This method follows the social hierarchy and hunting
Sha
behavior of grey wolves. Their hunting strategy is followed
encoder
PID Reference by this algorithm to search and hunt a prey (solution).
controller Current There are three main steps in hunting.
generator

1. Tracking, chasing and approaching the prey.

θ
2. Encircling and harassing the prey until it stops moving.
3. Attacking the prey.

Fig. 2  Overall block diagram of a PID controlled BLDC motor drive Just like the social hierarchy of grey wolves (to live in
groups), in this algorithm, four groups are defined; namely
Alpha (α), Beta (β), Delta (δ), and Omega (ω). During the
designing stage the social hierarchy of wolves is modeled.
popular algorithms which was developed by Kennedy, J. and Alpha is the fittest solution; following Beta and Delta as
Eberhart, R.C (1995) [13]. Its structure and approach to a the second and third best solutions. The rest of the solu-
problem follows the behavior of birds at the time of search- tions are least important and considered as Omega.
ing food, escaping from hunters or searching of mates [14,
15].
Particles are conceptual entities similar to birds which 4.1 Searching for Prey
fly through the search space [6]. Each particle has two state
variables, i.e. current velocity ­(Vi+1) and current position According to the position of the α, β and δ, grey wolves
­(Xi+1). At the beginning the population of particles (also search for prey. They diverge from each other to search
referred as swarm) are initialized. Similarly, the position the prey.
and velocity of each particle are initialized randomly. After
each iteration, the velocity and position of each particle are |A| > 1 (10)
updated using the following equations. The position of a
particle gives a trial solution for the search problem [16,
17].
4.2 Encircling the Prey
( ) ( )
vi (t + 1) = w.vi (t) + C1 .𝜙1 pi (t) − xi (t) + C2 .𝜙2 gi (t) − xi (t)
(8) To model the encircling behavior mathematically, follow-
ing equations are proposed.
xi (t + 1) = xi (t) + vi (t + 1) (9)
⃗ = ||C.
D ⃗X ⃗ ||
⃗ P (t) − X(t) (11)
i = 1, 2....n | |

Where ‘w’ is the weighted inertia which represents the ⃗ + 1) = X

X(t ⃗ P (t) − A.
⃗D ⃗ (12)
degree of change of direction. C ­ 1 and C­ 2 are positive con-
stants. Similarly, ‘ 𝜙1 ’ and ‘ 𝜙2 ’ are two random numbers Where,
selected from a range of [0, 1]. ‘pi’ is the best position of
ith particle and ‘gi’ is the best particle among the swarm ⃗ = 2⃗a.⃗r1 − a⃗
A (13)
[18–20]. The velocity of a particle is updated by using their
previous velocity, distance of their current position from ⃗ = 2.⃗r2
C (14)
own best position and best position of neighbors given by
(8). The new position is given by (9). t  = Current iteration.

13

958 Journal of Electrical Engineering & Technology (2021) 16:955–961

⃗ C
A, ⃗ → Coefficient Vectors.
XP → Position Vector of the prey.
⃗
⃗ → Position vector of wolf.
X
The parameter ‘a’ is decreased from 2 to 0 in order to
emphasize exploration and exploitation, respectively. The
position of a grey wolf changes with respect to the position
of prey. In this algorithm, the optimum solution (prey) is
achieved with the help of the till known three best solu-
tions (α, β, δ). To update their positions at next iteration,
the following formulae are used.

⃗ 𝛼 = ||C
D ⃗ .X ⃗ −X ⃗ ||
| 1 𝛼 |
⃗ 𝛽 = ||C
D ⃗ 2 .X ⃗ ||
⃗𝛽 − X (15)
| |
⃗ |⃗ ⃗ ⃗ |
D𝛿 = |C3 .X𝛿 − X |
| |
( )
⃗1 = X
X ⃗𝛼 − A ⃗𝛼
⃗ 1. D
( )
⃗2 = X
X ⃗𝛽 − A ⃗𝛽
⃗ 2. D (16)
( )
⃗3 = X
X ⃗𝛿 − A ⃗𝛿
⃗ 3. D

⃗1 + X
X ⃗2 + X
⃗3
⃗ + 1) =
X(t (17)
3 Before designing a optimization technique based PID
controller the objective function (also called fitness func-
4.3 Attacking the Prey tion) is first defined by taking the desired specifications
and constraints into consideration [22, 23]. A proper
The wolves converge towards the prey, i.e. the position of objective function is chosen to tune the controller param-
prey is the final position of Alpha. eters by considering entire closed loop response. There are
many time domain functions which can act as objective
|A| < 1 (18)
After each iteration, the GWO algorithm allows its search
agents to update their position based on the location of α, β,
δ and attack towards the prey.
Before starting the main objective of any meta-heuris-
tic population-based algorithm; two basic parameters are
required to be initialized. The first and foremost parameter
is the “maximum number of search agents”. In GWO algo-
rithm we recognize the search agents as “grey wolfs”. In
(a)
case of PSO the search agents are called as “swarm”. The
number of search agents may vary according to the applica-
tion. In this application, this value is taken as 30. The second
important parameter is the “number of iterations”. This also
depends upon the type of application and varies in a broad
range. The less the number of iterations; less the evaluation
time. The maximum number of iterations indicate, that the
program stops here whether the optimal solution is achieved (b)
or not. In this program this value is taken as 50. The pseudo
code of this algorithm is shown below. Fig. 3.  Simulation model of (a) BLDC motor, (b) PID controlled sys-
tem.

13
Journal of Electrical Engineering & Technology (2021) 16:955–961 959

Table 1  BLDC Motor parameters Table 2  Basic parameter values of the algorithm

BLDC Motor parameters Values Algorithmic parameters Values

RS 0.5 ohm Max. no. of search agents 30

L 8 mH iterations 50
J 0.0465 kg.m2
K 0.55 kg.m/A
b 0.004 N.m.sec/rad Table 3  PID controller parameters for different algorithms
Ia 10 A
Algorithm KP KI KD
V 12V
P 84W PSO 0.0707 0.0582 −0.0582
Ψm 65 mV/rad/sec GWO 0.0898 0.0605 0.0138
TP 2.9 N.m
p 8
GWO algorithm is implemented for optimizing the
objective function and after that performance of PSO and
function for different systems. These are broadly classi- GWO is compared.
fied into two groups. (a) Criteria based on a few points
in the response, (b) Criteria based on the entire response
or integral criteria. The integral criteria are generally 5 Implementation and Simulation Results
chosen because of its good performance. An extra advan-
tage of using the integral function is that it can be easily The simulation model for the purposed PID controlled
extended to a multi-loop system. The objective function BLDC motor drive is shown in Fig. 3. A reference speed
proposed for this system “integral of product between time is set as per the requirement. The measured speed is fed to
and absolute velocity error (ITAE) of the motor” [24] is the comparator via the feedback path. The error signal is
defined in (19). processed by the PID controller.
∞ The different parameters of the motor are given in
Table 1. The parameters (­ KP, ­KI, ­KD) of the PID controller
∫ (19)
ITAE = t|e(t)|dt
are evaluated by using different algorithms. Here we use
0 GWO algorithm to evaluate the PID controller parameters
and the output response is compared with the PSO based

Fig. 4  Output response of PSO 1800

and GWO based systems
1600

1400

1200
Speed (RPM)

1000

800

600

400
PSO
GWO
200

0
0 5 10 15
Time (Sec)

13

960 Journal of Electrical Engineering & Technology (2021) 16:955–961

Table 4  Performance of PSO and GWO based PID controllers 6 Conclusion

Algorithm Rise time (s) % Overshoot Settling time Peak time (s)
(s) PID controller is a cost-effective choice for a BLDC motor
drive [25]. To improve the performance of the PID controller
PSO 1.99 6 4.022 1.075
many advanced techniques are evolving. Some techniques
GWO 2.035 2.6 3.0 1.4
like PSO already proven its effectiveness in determining the
parameters of the PID controller. In this article a new opti-
mization technique (GWO) is applied which shows better
method. For both the algorithms the numbers of search results in controlling the speed of a BLDC motor than PSO
agents and iterations are shown in Table2. Similarly, the technique. From results it is seen that the machine is settled
values of K­ P, ­KI and K
­ D for both the methods are shown down faster compared to PSO based technique. Also, the
in Table 3. suggested method shows lesser damping. The stability analy-
By using the controller parameters as shown in the sis performance criterion viz ISE, IAE, ITAE, ITSE values
Table 3 the output responses are obtained for both the cases are much improved in this suggested technique. Though
as shown in Fig. 4. the rise time is slightly higher than PSO technique but the
From the Fig. 4 it is clear that the peak overshoot and other improvements in the time domain performance encour-
settling time is less in case of GWO based PID controlled age the usage of GWO technique to tune the PID control-
system than the PSO based system. The performance of both ler parameters of BLDC motor. The proposed method may
the systems is compared in Table 4 when the motor runs at give a new dimension towards the controller design field for
1500 rpm. Figure 5 shows how the solution was searched in a BLDC motor drive system.
the search space using GWO (Table 5).

Parameter space Objective space

GWO

25

20

1
10
F10( x , x )

15
2

Best score obtained so far

1

10

0
20

10 20
10
0
0
x
2 -10 x1
-10
-20 -20 5 10 15 20 25 30 35 40 45 50
Iteration

Fig. 5  GWO search space and pattern of search

13
Journal of Electrical Engineering & Technology (2021) 16:955–961 961

Table 5  Performance of PSO and GWO based PID controllers 16. Kennedy J. (2006). Swarm Intelligence. In: Zomaya A.Y. (eds)
Handbook of Nature-Inspired and Innovative Computing.
Algorithm ISE IAE ITAE ITSE Springer, Boston, MA. https​://doi.org/10.1007/0-387-27705​-6_6
17. Eberhart R, Shi Y, Kennedy J (2001) Swarm intelligence. Elsevier.
PSO 1233 53.43 1293 696.9 https​://doi.org/10.1016/B978-1-55860​-595-4.X5000​-1
GWO 1035 46.86 915.9 489.6 18. Gaing ZL (2004) A particle swarm optimization approach for opti-
mum design of PID controller in AVR system. Energy Convers
IEEE Trans 19(2):384–391
19. Panda S, Padhy N (2008) Comparison of particle swarm optimiza-
References tion and genetic algorithm for FACTS-based controller design. Int
J Appl Soft Comput 8(4):1418–1427
1. Suresh Kumar B, Varun Raj D, Venkateshwara Rao D (2020) 20. Zarringhalami M, Hakimi S, Javadi M (2010) Optimal regulation
Speed control of BLDC motor with PI controller and PWM tech- of STATCOM controllers and PSS parameters using hybrid parti-
nique for antenna’s positioner. In: Hemanth D, Kumar V, Malathi cle swarm optimization, IEEE conference. https:​ //doi.org/10.1109/
S, Castillo O, Patrut B (eds) Emerging trends in computing and ICHQP​.2010.56254​36
expert technology COMET lecture notes on data engineering and 21. Mirjalili Seyedali, Mirjalili Seyed Mohammad, Lewis Andrew
communications Technologies. Springer, Cham (2014) Grey wolf optimizer. Adv Eng Softw. https ​ : //doi.
2. Yedamale P (2003) Brushless DC (BLDC) Motor fundamentals. org/10.1016/j.adven​gsoft​.2013.12.007
Microchip Technology Inc. 20:3–15 22. Potnuru D, Tummala ASLV (2019) Grey wolf optimization-based
3. Bianchi N, Bolognani S, Jang JH, Sul SK (2006) Comparison improved closed-loop speed control for a BLDC motor drive. In:
of PM motorstructures and sensor less control techniques for Satapathy S, Bhateja V, Das S (eds) Smart intelligent comput-
zero-speed rotorposition detection. Power Electron IEEE Trans ing and applications smart innovation, systems and technologies.
22(6):2466–2475 Springer, Singapore
4. Dutta P, Mahato SN (2014) Design of mathematical model and 23. Muniraj M, Arulmozhiyal R, Kesavan D (2020) An improved
performance analysis of BLBLDC motor. IEEE Int Conf Con- self-tuning control mechanism for BLDC motor using grey wolf
trol Instrum Energy Commun India. https​://doi.org/10.1109/ optimization algorithm. In: Bindhu V, Chen J, Tavares J (eds)
CIEC.2014.69591​30 International conference on communication, computing and elec-
5. Fogel DB (2006) Evolutionary Computation: Toward a New Phi- tronics systems lecture notes in electrical engineering. Springer,
losophy of Machine Intelligence. Press Series on Computational Singapore
Intelligence IEEE, 3rd Edition: 128-138 24. Zitzler Eckart, Simon Künzli (2004) Indicator-based selection in
6. Ziegler G, Nichols NB (1942) Optimum settings for automatic multiobjective search parallel problem solving from nature-PPSN
controllers. Trans ASME 64:759–768 VIII. Springer, Berlin Heidelbergss
7. Zhang L (2004) “Simplex method based optimal design of PID 25. Shamseldin MA, EL-Samahy, AA (2014) Speed control of BLDC
controller.” Inform Control 33:376–379 motor by using PID control and self-tuning fuzzy PID control-
8. Thomas Neenu, Poongodi Dr P (2009) Position control of DC ler," 15th International workshop on research and education in
motor using genetic algorithm based PID controller. Proceedings mechatronics (REM), El Gouna pp. 1-9, https​://doi.org/10.1109/
of the World Congress on Engineering. 2 REM.2014.69204​43
9. Kim Dong Hwa, Jin Ill Park (2005) Intelligent PID controller
tuning of AVR system using GA and PSO. Springer-Verlag Ber- Publisher’s Note Springer Nature remains neutral with regard to
lin Heidelberg: ICIC, Part II, LNCS 3645: 366-375. https​://doi. jurisdictional claims in published maps and institutional affiliations.
org/10.1007/11538​35638​
10. Rubaai A, Kotaru R (2000) Online identification and control of a
DC motor using learning adaptation of neural networks. Ind Appl
IEEE Trans 36(3):935–942 Pallav Dutta  is working as an Assistant Professor at Aliah Univer-
11. Shyam A, Daya FJL (2013) A comparative study on the speed sity and presently serving as officiating Head of Electrical Engineer-
response of BLDC motor using conventional PI controller, anti- ing Department. Prior to Aliah University, he also served as a faculty
windup PI controller and fuzzy controller, International Confer- member in various reputed Govt. Universities/ Institutes in different
ence onControl Communication and Computing (ICCC): https​:// parts of India. He hasbeen engaged in teaching/research/administration
doi.org/10.1109/ICCC.2013.67316​26 for the past 12 years. He received the Bachelor’s degree in Electrical
12. K Ang, G Chong, Y Li (2005) PID control system analysis, design, Engineering from West Bengal University of Technology and the Mas-
and technology. Control System Technology IEEE Transaction on, ter’s degree in Electrical Engineering from University of Calcutta. His
13(4): 559 576 research interests include power electronicsand motor control systems.
13. Kennedy J, Eberhart RC (1995) Particle swarm optimization.
Proc. IEEE International Conference on Neural Networks,Perth, Santanu Kumar Nayak  was born in Khordha, Odisha, India. He
Australia, IEEE Service Center, Piscataway, NJ: 1942- 1948. received Bachelor of Technology Degree in Electrical and Electronics
https​://doi.org/10.1109/ICNN.1995.48896​8 Engineering from Trident Academy of Technology, Bhubaneswar, in
14. Shi Y, Eberhart RC (1998) A modified particle swarm optimizer. the year 2012 and completed Master of Technology in Power Electron-
IEEE International Conference on Evolutionary Computation, ics Control and Drives from Veer Surendra Sai University of Technol-
Anchorage, Alaska. https​://doi.org/10.1109/ICEC.1998.69914​6 ogy, Burla, Odisha. He is currently working in Infosys Ltd, Hyderabad.
15. Shi, Y, Eberhart RC (1999) “Empirical Study of Particle Swarm His research areas are BLDC motor drive and optimization techniques.
Optimization.” Proceedings IEEE: 1945 –1950. https​: //doi.
org/10.1109/CEC.1999.78551​1

13