International Journal of Electrical and Computer Engineering (IJECE)

Vol. 14, No. 6, December 2024, pp. 6225~6236

ISSN: 2088-8708, DOI: 10.11591/ijece.v14i6.pp6225-6236  6225

Performance enhancement of brushless direct current motor

under different novel optimization techniques

Babu Ashok, Mahesh Kumar

Department of Electrical and Electronics Engineering, Puducherry Technological University, Pondicherry, India

Article Info ABSTRACT

Article history: This research paper presents a novel attempt of speed control for brushless
direct current (BLDC) motor in low power/servo motor applications. The
Received Dec 12, 2023 performance is measured based on the swiftness for the recovery of desired
Revised Jul 20, 2024 speed amidst in disturbances, sensitive to supply/motor load fluctuations.
Accepted Aug 6, 2024 The proportional integral (PI) controller is competent only for linear time
invariant systems. The state of art technology is, PI controller is used with
metaheuristic optimization algorithms viz. Honeybee mating optimization
Keywords: (HBO), artificial immune system (AIS), and frog leaping guided algorithm
(FLG), for fine tuning of gain coefficients. Earlier literature survey shows
Artificial immune system power quality and time domain specifications for separate applications. An
Brushless direct current motor innovative approach for the assessment of performance indicators like
Frog leaping guided maximum overshoot (𝑀𝑝 ), settling time (𝑡𝑠 ), power factor (PF) and total
Honeybee mating optimization harmonic distortion (THD) simultaneously in the optimized PI controller is
Metaheuristic algorithm suggested. By avoiding local optima trapping, this method gives better
Speed controller dynamic performance for various test conditions. MATLAB/Simulink 2021a
software is utilized in the examination of performance in various load and
speed scenarios, subsequently validated with hardware where cost effective
Arduino controller replaced programmable interface controllers (PIC)
microcontroller.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Babu Ashok
Department of Electrical and Electronics Engineering, Puducherry Technological University
Pondicherry, India
Email: babuashok69@ptuniv.edu.in

1. INTRODUCTION
New developments in compact motors have recently resulted from the introduction of sophisticated
power electronic devices and contemporary control engineering. Their straightforward and economical
speed-controlling techniques and linear speed-torque characteristics have helped direct current motors
account for more than 70% of fractional horsepower motors used in the electrical sector [1]. However, direct
current (DC) motors have palpable drawbacks as well, like a shorter lifespan due to mechanical friction
across the brushes and commutator [2]. On the other hand, the permanent magnet synchronous motor
(PMSM) in this brushless direct current (BLDC) motor is electronically commutated. It is made up of three
armature coils in the fixed section and special magnets (rare earth) contained in the spinning part. Compared
to PMSM, which has a sinusoidal counter-electromagnetic force (counter-EMF) waveform structure, it has a
non-sinusoidal (counter-EMF) and 15% more power density [3]. Using an electronic regulator, BLDC motors
eliminate a common defect seen in many conventional electric drives, namely the mechanical commutator
(brushes). However, this restricts the motor's size and output. Therefore, for this motor drive circuit an
accurate and complex actuator controller was required.

Journal homepage: http://ijece.iaescore.com

6226  ISSN: 2088-8708

The strong performance and resilience of the BLDC market drive its remarkable expansion.
Prior to 2030, BLDC motors will replace other drives due to their increasing popularity [4]. Reduction in
electrical connections, mechanical misalignments, final product size, and weight, the development of BLDC
motors is expected to improve product durability and dependability. Automobiles, electric trains, robotics,
aerospace, home appliances, computer peripherals, the food and chemical industries, healthcare equipment,
and many more industrial applications are a few of the many uses for it [5].
The primary factors that have palpable influence on the tracking and regulation of BLDC motors
include controller design [6] and controller gain optimization [7]. The use of highly effective controllers of a
small number and innovative control strategies is seen in literature for a change in motor behavior, with focus
on either power quality or time domain specification as the research gap. This study is related to the
investigation of the effects of Time domain characteristics, viz., 𝑡𝑠, 𝑀𝑝 , 𝐸𝑠𝑠 , 𝑡𝑟 and power quality issues viz.,
power factor (PF) and total harmonic distortion (THD) simultaneously as a novel attempt. A variety of speed
controllers and state-of-the-art controlling techniques have been used for solving this problem . Metaheuristic
optimization is a unique technique that can handle uncertainties and nonlinear parameters. Hence it has been
used for the fine tuning of the controller gains considering its ability to provide appropriate solutions for the
problems increased in nonlinear programming-hard situations [8]. The three contributions to improve the
dynamic properties of BLDC drives are listed below.
− The artificial immune system (AIS), honeybee mating optimization (HBO) control system is intended for
analysis of the controller's performance in terms of several metrics related to motor dynamics.
− The frog leaping guided (FLG) algorithm, sometimes known as “Frog jumping algorithm” is a powerful
optimization technique in the drive circuits.
− Comparison with results seen in hardware, helps verification of the behavior of the motor with the use of
the simulated outcome.
The paper is organized as follows: section 2 deals with outline. Section 3 discusses speed controllers
that use different unique optimizations such as HBO, AIS, and FLG. Section 4 refers to results and
discussion; section 5 presents the hardware design and results. Section 6 reflects the conclusions to ensure
that it meets the benchmarks set by the IEEE standards by comparing with simulated results.

2. WORK-OUTLINE
Various metaheuristic algorithms with inspiration from nature have been developed recently for
maximation of the control system gains of BLDC motors. Smart computing techniques are becoming
increasingly popular in control engineering, as demonstrated [9]. It has been suggested the adaptive fuzzy
proportional–integral–derivative (PID) control speed for improvement in the speed control characteristic of the
BLDC drives [10]. An adaptive fuzzy logic controller (FLC) and a simple PID controller are components of
the suggested system. This method uses genetic algorithm (GA) for the fine turning of the controller gains for
the management of speed of the BLDC motor till reaches a specific steady state condition. A GA optimization
fuzzy-PID controller for improvement in behavior created [11], adaptive factor, multi-objective equation and
improved differential evolutional (DE) approach are used to optimize the gains of the controller [12].
Particle swarm optimization (PSO), is a method for increasing controller goals [13]. The study
demonstrated that, in comparison to GA-based controllers, the suggested PSO based algorithm produced
lower levels of overshoot (63.94%), peak time (0.26 seconds), and correction time (1.4 seconds). According
to these findings, choice of the controller's parameter has a major impact on performance. The authors used a
traditional approach with this algorithm for the application of drives. Using an enhanced DE approach,
Beig et al. [14] suggested switching pattern for pulse width modulation technique to control speed.
The successful application depends upon component identification in power converter based on
application [15]. This technique is used to achieve the minimizing error while maximizing controller gains.
Salp swarm optimization algorithm was proposed for BLDC motors with sensors [16]. In order to maximize
the benefits of a conventional controller, an adjustable speed and power factor correction controller is
proposed [17], however they did not analyze the THD. A grey wolf optimization technique was presented for
the optimization of the gains of the proportional–integral (PID) controller of BLDC motors [18].
Simultaneous analysis of the time domain specifications and power quality issues through optimization is the
need of the hour as shown in Figure 1. This will ensure higher impact in the drive system. IEEE standard
519-2014 recommends the harmonics limits, lower THD implies low electromagnetic interference (EMI),
low heating and low iron core loss. Stator voltage harmonics can generate torque ripple, which impedes
smooth operation and increases heat and in addition to that positive sequence component is responsible for
overheating and negative sequence component is responsible for torque ripple.

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236
Int J Elec & Comp Eng ISSN: 2088-8708  6227

Figure 1. Structure of controller

3. SPEED CONTROLLER USING VARIOUS OPTIMIZATIONS-SIMULATED RESULTS

3.1. Honeybee mating optimization (HBO tuned controller for BLDC motor)
A goal of optimization-based solutions in this digital era is to reduce manual labor and solution
time. Researchers have produced a plethora of software packages that achieve this dual benefit. Honey bee
mating optimization [19] is employed as a soft computing method where iterations are carried out until the
desired speed is reached.
Annealing function for queen probability, accept 𝑖 th drone to mate and improve trial solution is,
−∆(𝑓𝑖)
[ ]
𝑃 (𝑄𝑢𝑒𝑒𝑛, 𝐷𝑟𝑜𝑛𝑒) = 𝑒 𝑒𝑛𝑒𝑟𝑔𝑦(𝑡) (1)

∆(𝑓) refer, modulus value of fitness difference with drone and queen. When mating flight is in progress both
queen’s speed and energy reduce simultaneously after each iteration.

𝑒𝑛𝑒𝑟𝑔𝑦 (𝑡 + 1) = ∞ 𝑥 𝑒𝑛𝑒𝑟𝑔𝑦 (𝑡) (2)

where 𝑡 € [0, 1, 2, … 𝑡] and decay rate 0 < ∞ < 1

𝑠𝑝𝑒𝑒𝑑 (𝑡 + 1) = 𝑒𝑛𝑒𝑟𝑔𝑦 (𝑡) − 𝛽 𝑤ℎ𝑒𝑟𝑒 𝑡 € [0, 1, 2 … . 𝑡]

𝑎𝑛𝑑 𝑑𝑒𝑐𝑎𝑦 𝑟𝑎𝑡𝑒 𝛽 𝑤𝑖𝑡ℎ𝑖𝑛 [0, 1] (3)

Queen updates its energy and speed with the use of (2) and (3). Mating flight ends when the energy level falls
below a threshold value (close to zero).
− Queen: Highest weightage in the cluster at the current instant.
− Drone: Low weightage

3.2. Artificial immune system

Life becomes fatal in the absence of immune system. AIS mimics’ human immunology applied to
complicated issues. A single antigen can be recognized by several antibodies. As shown in Figure 2, AIS
distinguishes between antigen and antibody; this phenomenon is utilized as a tool for nonlinear and time
variant applications [20]–[22]. The following steps are involved in the process.
a. Initialization
Antigen-value to be optimized is objective function 𝑓(𝑥) and antibodies-corresponding solution to
the problems.
b. Cloning

𝐴𝑛𝑡𝑖𝑏𝑜𝑑𝑖𝑒𝑠 𝑎𝑟𝑒 𝑐𝑙𝑜𝑛𝑒𝑑 𝑏𝑎𝑠𝑒𝑑 𝑜𝑛 𝑡ℎ𝑒 𝑓𝑖𝑡𝑛𝑒𝑠𝑠, 𝑁𝐶 = Ʃ (𝛽 ∗ 𝑗/𝑖) (4)

Performance enhancement of brushless direct current motor under different … (Babu Ashok)
6228  ISSN: 2088-8708

where 𝑖 = 1, 2, 3 … 𝑛, 𝑁𝐶 is clone number, 𝛽 means multiplier factor, and 𝑗 refers antibodies population
size.
c. Hyper mutation
Clones are mutated in inverse proportion of affinity and N antibodies-selected for the next iteration.
If antibodies match antigen (threshold value) then concentration increases ‘stimulation’ if not
concentration decreases ‘suppression’.
d. Repeat
Introduce a random number till antibodies are generated.
e. End
Stopping criteria met, i.e., antibody concentration will nullify antigen.

Figure 2. Antigen nullified by antibody

3.2. Frog leaping with global guided algorithm

This study has developed a metaheuristic method called frog leaping with global guided
optimization (FLG) meant for a change in gain coefficients for control. The global best frog will be replaced
by worst frog for the creation of FLG, the global best frog is used in place of the worst frog. Memes in FLG
therefore spread at a quicker rate of convergence. The optimal values of the scaling factors, namely
𝐾𝑝 and 𝐾𝑖 of the controller, are ascertained using the proposed FLG. This approach has been verified through
simulations and contrasted with real-world BLDC drive implementations.
A swarm of frogs has been used in FLG's population-based, evolving metaheuristic, which
maximizes prey [23]–[25]. Its integration of a memeplex algorithm based on genetic evolution with particle
swarm optimization (PSO) is a major factor in its success. The PSO algorithm serves as the foundation for
this method's local exploration, and the idea is integration of data from concurrent local searches originates
from the shuffling complex evolution methodology. Frog population shows a cluster of potential solutions in
FLG. The frog sets were divided into several mimetics’ groups, each of which represents an exclusive
temperament. The frogs tend to congregate at what appears to be the best; this is just provisional and could be
a local optimum. However, some memetics were taken in sub-memetics to avoid entrapment in local optima.
It is best to move the frog that is at the worst place. A new group known as memeticists has been developed
throughout memeplex iterations. FLG steps involved are.
a. Population creation: The initial population (frogs) is defined as

𝑃 = {𝑝𝑖 , 𝑓𝑖 , 𝑖 = 1,2,3 … . 𝐹} 𝑤𝑖𝑡ℎ𝑖𝑛 𝑡ℎ𝑒 𝑒𝑥𝑝𝑙𝑜𝑟𝑎𝑡𝑖𝑜𝑛 𝑎𝑟𝑒𝑎,

where 𝑝𝑖 is the position of 𝑖 th frog and 𝑓𝑖 represents its fitness. The fitness values are arranged in
decreasing order.
b. Splitting the memetics: Divide the population into 𝑛 memeticists {𝑄1 , 𝑄2 , 𝑄3 … . . 𝑄𝑛 }, each contains 𝑛
frogs and

𝑄𝑖 = [(𝑝𝑗 , 𝑓𝑗 )] | 𝑝𝑗 = 𝑝𝑗+𝑛(𝑗−1) , 𝑓𝑗 = 𝑓𝑗+𝑛(𝑗−1) (5)

where 𝑗 = 1,2,3 … . . 𝑚.
c. Submemetics creation: The strategy selection of a submemetics (fork frogs) in every memeticists has
larger parameters distributed in good locations. Better position in frog strategy has greater weights
allocated to submemetics having triangular probability distribution as (6).

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236
Int J Elec & Comp Eng ISSN: 2088-8708  6229

2(𝑚+1−𝑖)
𝑤𝑖 = , 𝑖 = 1,2,3 … , 𝑚 (6)
𝑚(𝑚+1)

d. Submemetics evolution: Let𝑝𝑏𝑒𝑠𝑡 is the best location and 𝑝𝑤𝑜𝑟𝑠𝑡 is the worst location in submemetics.
Then, local exploration starts from the worst frog to leap in best group. The current position is updated by
one leaping step as shown in (7).
g g g g
min{int[r g (pbest − pworst )], lmax } , if pbest ≥ pworst
lg+1 = { g g g g (7)
min{int[r g (pbest − pworst )], −lmax } , if pbest < pworst

where 𝑖𝑛𝑡(. ) denotes the integer function which converts the specified value into an integer number;
min(. ) function returns the item with the lowest value in an iterable; 𝑙𝑚𝑎𝑥 is maximum leap size;
𝑟- random number and 𝑔-evolution generation. Moreover, if the current location is better than worst frog,
then worst frog’s position is modified as
𝑔+1 𝑔
𝑝𝑡 = 𝑝𝑡 + 𝑙 𝑔+1 (8)

Else, the worst frog position becomes global optima hence

𝑔 𝑔 𝑔 𝑔
𝑚𝑖𝑛{𝑖𝑛𝑡[𝑟 𝑔 (𝑝𝑠𝑏𝑒𝑠𝑡 − 𝑝𝑤𝑜𝑟𝑠𝑡 )], 𝑙𝑚𝑎𝑥 } , 𝑖𝑓 𝑝𝑠𝑏𝑒𝑠𝑡 ≥ 𝑝𝑤𝑜𝑟𝑠𝑡
𝑙 𝑔+1 = { 𝑔 𝑔 𝑔 𝑔 (9)
𝑚𝑖𝑛{𝑖𝑛𝑡[𝑟 𝑔 (𝑝𝑠𝑏𝑒𝑠𝑡 − 𝑝𝑤𝑜𝑟𝑠𝑡 )], −𝑙𝑚𝑎𝑥 } , 𝑖𝑓 𝑝𝑠𝑏𝑒𝑠𝑡 < 𝑝𝑤𝑜𝑟𝑠𝑡
𝑔
where 𝑝𝑠𝑏𝑒𝑠𝑡 is best position of the swarm. In case if the worst frog fails to improve its position, a random
position is automatically generated to replace it
𝑔+1
𝑝𝑡 = 𝑏1 + 𝑖𝑛𝑡[𝑟 𝑔 (𝑏1 − 𝑏2 )] (10)

where [𝑏1 , 𝑏2 ] - boundary for possible location of frogs. Then, frogs are sorted in decreasing order,
ascertained on eligibility. Above steps are repeated till submemetics 𝐺1 is formed.
e. The remaining memetics have been rearranged in the decreasing order of fitness on completion of this
local investigation. This is known as memeticists shuffling. Until memetic evolution generation 𝐺2 is
reached, the group is divided into memeticists, and the local exploration process is carried out repeatedly.

4. RESULTS WITH DISCUSSION

Using MATLAB/Simulink 2021 software, the BLDC drive circuits were designed for the
verification of the speed of this innovative approach. Using the FLG technique, the effectiveness of speed
control for BLDC drives is confirmed under various test scenarios. The AIS and HBO are seen as two further
sophisticated control systems with the performance of comparison with FLG controller. All the three
algorithms were utilized until the results were obtained. Three distinct instances have been taken into
consideration for analysis of the performance of the suggested control system: i) speed response for constant
load, ii) speed response for step change in load, and iii) response for variable speed. In optimizations the
computation time of algorithm is proportional to constraints, it proceeds iteratively selecting a new solution
until predefined halt condition is met.

4.1. Speed response for constant load

4.1.1. Simulation results
BLDC motors are designed for continuous duty. Interaction of the magnetic forces of permanent
magnet (rotor) with electromagnetic field of armature (Stator) is seen in the production of torque. The load
torque during a load test which should be identical to the operating torque, is determined by the duty cycle
(d) which the controller generates based on the speed error, as illustrated in the block diagram Figure 1
below. For different optimization strategies, Simulation predicts performance for different optimization
strategies with the use of static loading at 25%, 50%, 75%, and 100% of full load. Table 1 and Figures 3 to 5
suggest that FLG is strong robustness, improved convergence speed than to HBO and AIS. The algorithm
should choose “d” as the ideal value considering increasing the switching frequency will result in more pulse
width modulation (PWM) losses and decreasing it will result in less system bandwidth, which could harm the
drive system or cause stoppage of the motor.

Performance enhancement of brushless direct current motor under different … (Babu Ashok)
6230  ISSN: 2088-8708

Table 1. Performances for different loading @ 2000rpm

Method Performance metrics 25% 50% 75% 100%
I=0.3A I=0.6 I=0.9 I=1.2A
HBO THD 15.94 16.2 18.22 20.71
𝑀𝑝 80 140 210 260
𝑇𝑠 0.6 0.8 0.9 1
PF 0.9364 0.9274 0.921 0.9142
AIS THD 13.28 14.3 15.7 17.25
𝑀𝑝 40 100 140 160
𝑇𝑠 0.5 0.16 0.2 0.25
PF 0.9586 0.951 0.9475 0.9361
FLG THD 7.67 11.96 12.4 13.62
𝑀𝑝 30 80 120 150
𝑇𝑠 0.09 0.1 0.15 0.2
PF 0.988 0.98 0.974 0.971

Figure 3. Speed response of BLDC motor @100% loading

Figure 4. THD comparison for static loading Figure 5. PF comparison for static loading

4.2. Response for step change in load using optimizations

As shown in Figure 6 and Table 2, FLG performs better than other methods for different step
changes in loading under different optimizations. Induction or electromechanical energy conversion is the
same operational principle that underpins all electrical motors, regardless of their various kinds and sizes.
Fast Fourier transform (FFT) analysis in MATLAB is used in the measurement of THD (IEEE standard
519-2014 recommends the harmonics limits, lower THD implies low EMI, low heating and low iron core
loss in motors), done in addition to PF in BLDC motors, considering the menace of THD for the overheating

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236
Int J Elec & Comp Eng ISSN: 2088-8708  6231

of windings that trips relays. Step changes in loading, such as pressing, cutting, and drilling, can benefit from
the step change in load analysis.

Figure 6. Speed responses for step change in load using optimization techniques

Table 2. Performances for step change in load

Method Metrics 25-0% 0-75% 75-50% 50-100%
HBO 𝑀𝑝 80 220 75 200
𝑇𝑠 0.50 0.70 0.60 0.65
PF 0.961 0.952 0.948 0.945
AIS 𝑀𝑝 75 210 55 120
𝑇𝑠 0.30 0.40 0.30 0.45
PF 0.973 0.965 0.917 0.957
FLG 𝑀𝑝 70 110 50 100
𝑇𝑠 0.20 0.30 0.20 0.40
PF 0.989 0.973 0.990 0.971

4.3. Response for variable speed

This research is most helpful for getting an understanding of the dynamics of an electric train during
its three phases of operation: acceleration, running, and braking. The goal is to provide passengers with a
comfortable ride while causing the least amount of wear and tear on the moving parts. The simulation's
output and waveforms shown in Figure 7 and Table 3 demonstrate the outperformance of the FLG-sponsored
optimization method with the other two methods in terms of 𝑇𝑠 , 𝑀𝑝 , PF, and THD.

Figure 7. Speed responses for variable speed @ constant (FULL) load using optimization techniques
Performance enhancement of brushless direct current motor under different … (Babu Ashok)
6232  ISSN: 2088-8708

Table 3. Speed change @ constant (full) load

Method Performance Metrics 0 to 1,500 1,500 to 1,250 1,250 to 1,000 1,000 to 1,250 1,250 to 15,00
HBO 𝑀𝑝 225 100 90 125 120
𝑇𝑠 0.7 0.5 0.5 0.6 0.6
PF 0.958 0.962 0.970 0.973 0.965
THD 20.71 19.63 17.88 19.12 20.67
AIS 𝑀𝑝 200 10 10 100 115
𝑇𝑠 0.3 0.15 0.15 0.25 0.15
PF 0.968 0.973 0.987 0.978 0.962
THD 17.25 16.61 15.16 16.49 16.85
FLG 𝑀𝑝 150 10 08 50 110
𝑇𝑠 0.2 0.10 0.10 0.15 0.10
PF 0.971 0.984 0.992 0.981 0.974
THD 13.62 12.57 11.98 12.05 13.03

4.3.1. Hardware results

The results of the simulation were verified using a hardware test. The two main forms of optimization
algorithms in the modern digital age were seen as deterministic and stochastic. Optimization is widely used for
addressing complicated problems. Although deterministic algorithms and stochastic algorithms, which deal with
the brainchild HBO, AIS, and FLG in this drive circuit, can be easily implemented by maintaining actual
dynamics as shown in Figure 8, Tables 4 and 5, the deterministic algorithm's objective function requires certain
constraints or assumptions, whereas the stochastic algorithm is not confirmed. Discrete values make BLDC
motor drive a complicated function for building, with the requirement of unique control strategy. While
commutating currents are delayed, a decrease in resistance provides a boost to the stator's electrical time
constant and steady state current. Increasing resistance results in more loss and a decrease in efficiency, it
requires attention for the determination of appropriate speed management. A 2% tolerance for speed variation
has been taken into consideration as per the international standards.

Figure 8. Snapshot of real time laboratory model hardware setup

Table 4. Hardware results @1500 RPM

Methods Performance metrics Loading
25% 50% 75% 100%
HBO IL 3.4 6.8 10.2 13.5
THD 28.9 36.2 42.3 48.4
PF 0.8912 0.8153 0.7821 0.6851
Speed error 90 130 195 290
AIS IL 3.4 6.8 10.2 13.5
THD 18.6 22.5 25.8 27.3
PF 0.9120 0.8263 0.7912 0.7125
Speed error 80 120 180 250
FLG IL 3.4 6.8 10.2 13.5
THD 8.3 12.2 15.4 17.6
PF 0.9812 0.9125 0.9025 0.9012
Speed error 7 12 16 18

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236
Int J Elec & Comp Eng ISSN: 2088-8708  6233

Table 5. Comparisons of simulated results with hardware for optimizations, Set speed=1500 rpm
Method Performance metrics Loading
25% 50% 75% 100%
Simln. Hardwr. Simln. Hardwr. Simln. Hardwr. Simln. Hardwr.
HBO THD 15.94 28.9 16.20 36.2 18.22 42.3 20.71 48.4
PF 0.9364 0.8912 0.9274 0.8153 0.9210 0.7821 0.9140 0.6851
Speed error 39.9 90 99.9 130 150 200 219.9 289.9
AIS THD 13.28 18.6 14.30 22.5 15.70 25.8 17.25 27.3
PF 0.9586 0.9120 0.9510 0.8263 0.9475 0.7912 0.9361 0.7125
Speed error 19.95 79.5 90 120 109.95 180 199.95 249.9
FLG THD 07.67 08.3 11.96 12.2 12.40 15.4 13.62 17.6
PF 0.988 0.9812 0.980 0.9125 0.974 0.9025 0.971 0.9012
Speed error 1.95 6.9 4.95 12 9.9 15.9 15 18

5. HARDWARE DESIGN
Coupling a motor with small value of mechanical time constant with large inertial load, shall result
in losing the merit of having a small moment of inertia. Whereas, when the motor with a large moment of
inertia if used for driving light load, the motor efficiency will be reduced. The most important feature of
BLDC motor is its ability to balance the power converter and load requirements through electronic
commutation. An experimental prototype of the proposed controller is shown in Figure 8. A bridgeless buck
boost converter intended for operation in discontinuous conduction mode (DCM) either inductor current or
capacitor voltage is discontinuous which is a prerequisite for the design [17] of drive circuits and in addition
to that to avoid gear reducer, coupling, and pulley, power converter is activated by controller.
a. Duty ratio calculation
The motor power rating is 750 W and the converter power is 850 W. Let the supply voltage
be 220 V root mean square (RMS) value, then input voltage is

2√2𝑉𝑠 2√2×220
𝑉𝑖𝑛 = = = 198 𝑉 (11)
𝜋 𝜋

𝑉𝑑𝑐
Voltage 𝑐𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛 ratio, 𝑑 = (12)
𝑉𝑑𝑐 +𝑉𝑖𝑛

Let the voltage limits across dc link control be

𝑉𝑑𝑐(𝑚𝑖𝑛), = 50 𝑉, 𝑉𝑑𝑐 (𝑚𝑎𝑥) = 200 𝑉

and 𝑉𝑑𝑐 (𝑛𝑜𝑚) = 100 𝑉 and the corresponding duty ratio 𝑑𝑚𝑖𝑛 = 0.2017 and 𝑑𝑚𝑎𝑥 = 0.5024 respectively.
b. Input inductors (𝐿1 and 𝐿2 )–design

𝑅(1−𝑑)2
𝐿𝑐 = (13)
2𝑓𝑠

𝐿𝑐 is calculated at 𝑑𝑚𝑖𝑛 the voltage is

𝑉𝑑𝑐(𝑚𝑖𝑛) = 50 𝑉 the corresponding power is 190 watts

2
𝑉𝑑𝑐(𝑚𝑖𝑛) (1−𝑑)2 502 (1−0.2016)2
𝐿𝑐 = × = × = 209.69 µH (14)
𝑃𝑚𝑖𝑛 2𝑓𝑠 190 2 𝑥 20000

𝐿𝑐
𝐿1 is taken as ,
10

Hence 𝐿1 = 𝐿2 = 25 µ𝐻 (15)

This size, weight and cost of the buck boost converter (BBC) is reduced.
c. DC link capacitor, 𝐶𝑑 – design
𝑃𝑜
𝐼𝑑 ⁄𝑉𝑑𝑐(𝑛𝑜𝑚) 850⁄
100
𝐶𝑑 = = = (16)
2𝜔 ∆𝑉𝑑𝑐 2𝜔 ∆𝑉𝑑𝑐 2×314×0.03×100

Performance enhancement of brushless direct current motor under different … (Babu Ashok)
6234  ISSN: 2088-8708

Assuming permissible voltage ripple across the dc link voltage of

3% = ∆𝑉𝑑𝑐 = 4511.67 µF (17)

Hence the nearest value is 5000 µF.

d. Filter circuit
A second order low pass inductance capacitance (LC) filter is applied across the supply for the
elimination of negative and zero sequence components as well as higher order harmonics.

𝐼𝑝𝑒𝑎𝑘 850⁄
220
𝐶𝑓 = 𝑡𝑎𝑛(Ф) = 𝑡𝑎𝑛(1) (18)
𝜔𝑉𝑝𝑒𝑎𝑘 314 𝑥 √2 𝑥 220
= 690.32 𝑛𝐹

𝐿𝑓 = 𝐿𝑟𝑒𝑞 + 𝐿𝑠 (19)

Let Ls = 4% of base impedance

1 1 Vs2
𝐿𝑓 = + 0.04 ( ) ( ) (20)
4π2 fc2 C f ω Po

1 0.04 2202
= + ( )
4π2 × 20002 × 690.32 × 10−2 314 850
= 0.00917347032 + 0.00725365305 = 0.0164
= 16 𝑚𝐻 (21)

5.1. Experimental setup

Hardware results detailed in Table 4 is compared with simulation results vide Table 5. Verification
of the outcomes of the simulation for the motor parameters is mentioned in Table 6. HBO seen having
lengthy search duration could get stuck in load optima details are seen in both simulation and hardware.
Recognition of alien things helped the immune system, sometimes referred to as “the second brain” in the
creation of self/non-self, non-linear networks from various antibodies. This can utilize the immunological
law for control and eradication of antigen. Any disruption is also eliminated by the software solution, which
uses digital filtering, instruction redundancy, software delay/reset and smoothing reactor/filter in hardware
for the regulation of BLDC drives.

Table 6. BLDC motor parameters (for Hardware)

Parameters Value
Number of poles 4
Power 750 Watts
Voltage 48 Volts
Current 13.5 Amp
Speed 1,500 rpm
Torque 18 Nm
Peak torque 24 Nm
Torque constant, KT 0.86
Rotor inertia 2.5 Kg

6. CONCLUSION
The controller optimization process and controller design are critical to the tracking and controlling
performance of BLDC motors. Hence proportional and integral gain values are predetermined for
conventional fixed gain methods, independent of test conditions, good results cannot be obtained for all
operational modes. Additionally, the outdated method does not guarantee working for one or more parameter
indicators such as peak overshoot, settling time, power consumption, steady-state accuracy. The drive circuit
requires the employment of many metaheuristic optimization methods, including HBO, AIS, and FLG, for
the attainment of the exact speed control. Effective fine tuning for the speed control was taken as the
objective for the achievement of good performance.

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236
Int J Elec & Comp Eng ISSN: 2088-8708  6235

Frog leaping algorithm with global guided principle (FLG), a unique optimization technique is
concluded as the best compared to other techniques with a minimum of 5% to 10% improvement in each
performance metrics. MATLAB/Simulink 2021 software is used for creation of the fine-tuned controller and
analysis of its performance under various load and speed situations. Total harmonic distortion, maximum
overshoot, settling time, power factor, and other performance metrics are used for assessment of the
efficiency of the controller.
Based on extensive empirical results, it is suggested that the optimization technique improves the
dynamic performance of the BLDC motor under a range of operating conditions. It is verified using an
Arduino controller in a real-time hardware experimental setup, closely matching simulated results, and
ensuring best-in-class safety in the motion control domain. This opens up new possibilities for global optima
for BLDC motor speed control.
Based on research findings it can be concluded that the proposed novel optimization technique
improved the BLDC motor dynamic performance under a range of operating conditions. Thus, obtained
results are validated with earlier research outcomes which are earmarked by IEEE standards. Simultaneous
analysis of time domain specifications and power quality indices are suitable for aerospace applications
where maintenance is not feasible.

REFERENCES
[1] Y. Lu, “DC motor control technology based on multisensor information fusion,” Computational Intelligence and Neuroscience,
vol. 2022, pp. 1–10, Jul. 2022, doi: 10.1155/2022/1447333.
[2] H. Y. Lee, S. Y. Yoon, S. O. Kwon, J. Y. Shin, S. H. Park, and M. S. Lim, “A study on a slotless brushless DC motor with
toroidal winding,” Processes, vol. 9, no. 11, Oct. 2021, doi: 10.3390/pr9111881.
[3] D. Mohanraj et al., “A review of BLDC motor: state of art, advanced control techniques, and applications,” IEEE Access, vol. 10,
pp. 54833–54869, 2022, doi: 10.1109/ACCESS.2022.3175011.
[4] H. Jigang, F. Hui, and W. Jie, “A PI controller optimized with modified differential evolution algorithm for speed control of
BLDC motor,” Automatika, vol. 60, no. 2, pp. 135–148, Apr. 2019, doi: 10.1080/00051144.2019.1596014.
[5] S. B. Joseph, E. G. Dada, A. Abidemi, D. O. Oyewola, and B. M. Khammas, “Metaheuristic algorithms for PID controller
parameters tuning: review, approaches and open problems,” Heliyon, vol. 8, no. 5, May 2022, doi:
10.1016/j.heliyon.2022.e09399.
[6] T. Wang, H. Wang, H. Hu, X. Lu, and S. Zhao, “An adaptive fuzzy PID controller for speed control of brushless direct current
motor,” SN Applied Sciences, vol. 4, no. 3, Feb. 2022, doi: 10.1007/s42452-022-04957-6.
[7] O. O. Martins, A. A. Adekunle, M. O. Arowolo, D. C. Uguru-Okorie, and B. O. Bolaji, “The effect of an evolutionary algorithm’s
rapid convergence on improving DC motor response using a PID controller,” Scientific African, vol. 17, Sep. 2022, doi:
10.1016/j.sciaf.2022.e01327.
[8] Suneeta, R. Srinivasan, and R. Sagar, “FPGA based control method for three phase BLDC motor,” International Journal of
Electrical and Computer Engineering, vol. 6, no. 4, pp. 1434–1440, Aug. 2016, doi: 10.11591/ijece.v6i4.10103.
[9] A. Rodríguez-Molina, M. G. Villarreal-Cervantes, O. Serrano-Pérez, J. Solís-Romero, and R. Silva-Ortigoza, “Optimal tuning of
the speed control for brushless DC motor based on chaotic online differential evolution,” Mathematics, vol. 10, no. 12, p. 1977,
Jun. 2022, doi: 10.3390/math10121977.
[10] R. Majdoubi, L. Masmoudi, M. Bakhti, A. Elharif, and B. Jabri, “Parameters estimation of BLDC motor based on physical
approach and weighted recursive least square algorithm,” International Journal of Electrical and Computer Engineering, vol. 11,
no. 1, pp. 133–145, Feb. 2021, doi: 10.11591/ijece.v11i1.pp133-145.
[11] S. Zhao, F. Blaabjerg, and H. Wang, “An overview of artificial intelligence applications for power electronics,” IEEE
Transactions on Power Electronics, vol. 36, no. 4, pp. 4633–4658, Apr. 2021, doi: 10.1109/TPEL.2020.3024914.
[12] Y. I. M. Al Mashhadany, A. K. Abbas, and S. S. Algburi, “Modeling and analysis of brushless DC motor system based on
intelligent controllers,” Bulletin of Electrical Engineering and Informatics, vol. 11, no. 6, pp. 2995–3003, Dec. 2022, doi:
10.11591/eei.v11i6.4365.
[13] G. Tang, P. Lu, X. Hu, and S. Men, “Control system research in wave compensation based on particle swarm optimization,”
Scientific Reports, vol. 11, no. 1, Jul. 2021, doi: 10.1038/s41598-021-93973-4.
[14] A. R. Beig, S. Kanukollu, K. Al Hosani, and A. Dekka, “Space-vector-based synchronized three-level discontinuous PWM for
medium-voltage high-power VSI,” IEEE Transactions on Industrial Electronics, vol. 61, no. 8, pp. 3891–3901, Aug. 2014, doi:
10.1109/TIE.2013.2288194.
[15] G. G. R. Sekliar and B. Baiiakara, “An internal current controlled BLDC motor drive supplied with PY fed high voltage gain DC-
DC converter,” International Journal of Electrical and Computer Engineering, vol. 8, no. 2, pp. 1262–1272, Apr. 2018, doi:
10.11591/IJECE.V8I2.PP1262-1272.
[16] W. N. Al-Din Abed, O. A. Imran, and A. N. Abdullah, “Sensored speed control of brushless DC motor based salp swarm
algorithm,” International Journal of Electrical and Computer Engineering, vol. 12, no. 5, pp. 4832–4840, Oct. 2022, doi:
10.11591/ijece.v12i5.pp4832-4840.
[17] V. Bist and B. Singh, “An adjustable-speed PFC bridgeless buck-boost converter-fed BLDC motor drive,” IEEE Transactions on
Industrial Electronics, vol. 61, no. 6, pp. 2665–2677, Jun. 2014, doi: 10.1109/TIE.2013.2274424.
[18] M. Varan, A. Erduman, and F. Menevşeoğlu, “A grey wolf optimization algorithm-based optimal reactive power dispatch with
wind-integrated power systems,” Energies, vol. 16, no. 13, Jun. 2023, doi: 10.3390/en16135021.
[19] R. Babu Ashok and B. Mahesh Kumar, “IoT monitored solar fed BLDC motor using landsman converter under partial shading
condition with hardware validation through HBMO algorithm,” Journal of Advanced Research in Dynamical and Control
Systems, vol. 11, no. 6, pp. 870–881, 2019.
[20] J. Timmis, “Artificial immune systems - Today and tomorrow,” Natural Computing, vol. 6, no. 1, pp. 1–18, Dec. 2007, doi:
10.1007/s11047-006-9029-1.
[21] A. Makhbouche, B. Boudjehem, I. Birs, and C. I. Muresan, “Fractional-order PID controller based on immune feedback
mechanism for time-delay systems,” Fractal and Fractional, vol. 7, no. 1, p. 53, Jan. 2023, doi: 10.3390/fractalfract7010053.

Performance enhancement of brushless direct current motor under different … (Babu Ashok)
6236  ISSN: 2088-8708

[22] I. Mahdavi, S. Firouzian, M. M. Paydar, and M. Saadat, “Simulated annealing and artificial immune system algorithms for cell
formation with part family clustering,” Journal of Industrial Engineering and Management Studies, vol. 7, no. 1, pp. 191–219,
2020.
[23] M. Eusuff, K. Lansey, and F. Pasha, “Shuffled frog-leaping algorithm: a memetic meta-heuristic for discrete optimization,”
Engineering Optimization, vol. 38, no. 2, pp. 129–154, Mar. 2006, doi: 10.1080/03052150500384759.
[24] B. Liang, J. Jiang, Z. Zhen, and K. Ma, “Improved shuffled frog leaping algorithm optimizing integral separated PID control for
unmanned hypersonic vehicle,” Transactions of Nanjing University of Aeronautics and Astronautics, vol. 32, no. 1, pp. 110–114,
2015.
[25] M. Mao, L. Zhang, P. Duan, Q. Duan, and M. Yang, “Grid-connected modular PV-Converter system with shuffled frog leaping
algorithm based DMPPT controller,” Energy, vol. 143, pp. 181–190, Jan. 2018, doi: 10.1016/j.energy.2017.10.099.

BIOGRAPHIES OF AUTHORS

Babu Ashok is currently working as Head of Department in Electrical and

Electronics Engineering of Karaikal Polytechnic College (Constituent college of PIPMATE –
Government of Puducherry) Karaikal. He is currently doing his Ph.D. at Puducherry
Technological University (erstwhile Pondicherry Engineering College), Pondicherry. He can be
contacted at e-mail: babuashok69@ptuniv.edu.in.

Mahesh Kumar has received the B.Tech. degree in electrical engineering from
Madurai Kamaraj University, MTech., in process control and instrumentation from N.I.T –
Trichy and Ph.D. from Jadavpur University, India. At present he is working as Professor in
electrical and electronics engineering of Puducherry Technological University (erstwhile
Pondicherry Engineering College) Pondicherry, South India and having several years of
teaching experience. He can be contacted at e-mail: bmk@ptuniv.edu.in.

Int J Elec & Comp Eng, Vol. 14, No. 6, December 2024: 6225-6236