2024 IEEE 3rd International Conference on Electrical Power and Energy Systems (ICEPES)

MANIT Bhopal (MP), India. Jun 21-22, 2024

Development of Robust Multi-Objective Mayfly

Optimization Method for Passive Power Filter
Design to Improve Power Quality
Pavankumar Daramukkala Kanungo Barada Mohanty
Dept. of Electrical Engineering Dept. of Electrical Engineering
National Institute of Technology National Institute of Technology
Rourkela, India Rourkela, India
[email protected] [email protected]
2024 IEEE 3rd International Conference on Electrical Power and Energy Systems (ICEPES) | 979-8-3503-9072-8/24/$31.00 ©2024 IEEE | DOI: 10.1109/ICEPES60647.2024.10653508

Abstract— Uncertain loads, especially nonlinear loads at in tuning the harmonic frequencies. The tuning of the PPF
the utility end of the electrical distribution network, create parameters depends mainly on the order of the harmonics to
harmonics in supply, in addition to the reactive power demand, be mitigated and the amount of reactive power to be
which is a significant issue as they deteriorate the power compensated simultaneously. Theoretically designed
quality at the utility side of the electrical power system in many parameters may not guarantee the exact tuning and other
aspects. This paper presents the implementation of a robust issues to get resolved completely as there are practical
multi-objective mayfly optimization (RMOMO) method for the limitations and effects that cause the PPFs to give
design of single-tuned passive power filters (PPFs) for power unsatisfactory performance practically when employed into
quality improvement. Reduction of the total harmonic
the system. Hence, optimizing the parameters with major
distortion (THD), improvement of the power factor, reduction
of losses, and effective cost management are the major
constraints included can give better solutions to the issues
objectives considered for the optimization problem, provided and improve the filter performance. Several optimization
with the reactive power compensation. Simulations are carried methods are evolving at a rapid pace as their applicability in
out using MATLAB/Simulink software, a comparative analysis diverse fields is showing proven and promising results with
is carried out, and the results are presented. The proposed more accuracy and fast response. Each optimization method
RMOO method excels many existing optimization methods in has its own unique feature of optimizing process and
terms of convergence, computational complexity, and response characteristics based on the objective functions
robustness. defined and constraints applied. Application of some of the
optimization methods, namely Teaching-Learning and
Keywords—Power quality, harmonics, passive power filter, Adaptive sine-cosine Moth-Flame algorithms, to hybrid
optimization, reactive power compensation. active power filters (HAPFs), is presented in [2] and [3],
whereas an improved Grasshopper optimization method and
I. INTRODUCTION
improved salp swarm algorithms are applied for the design
Increased usage of nonlinear loads such as switched- of HAPFs in [4] and [5]. Harmonics in the distributed
mode power supplies (SMPS), variable frequency drives systems are mitigated reportedly in [6] using the Harris
(VFDs), uninterruptible power supplies (UPSs), electronic hawks optimization method, whereas the single-tuned PPFs
ballasts, and power electronic converters cause the are designed in [7] using a multi-objective optimization
introduction of harmonics and other distortions into the method, namely the MIDACO algorithm. Another multi-
electrical system, particularly at the utility end of the objective method, namely the Pareto-Based Firefly
distribution system. The adverse effects of these nonlinear algorithm, is applied for PPF design in [8]. The allocation
loads include many power quality issues, such as overheating and sizing of PPFs are emphasized in [9], whereas the
and losses, voltage distortion, vibration, noise, unwanted hosting capacity of the distributed generated systems is
operation of protection equipment, and metering errors. addressed with the use of the optimal design of PPFs in [10].
Power filters are the traditional and effective solution to Multi-criteria are considered in [11] for the planning of PPFs
mitigate the harmonics [1]. They are broadly categorized as using a bacterial foraging optimization approach. Reactive
passive power filters (PPFs), active power filters (APFs), and power sharing is addressed in [12] among multi-arm PPFs.
hybrid power filters (HPFs). Each type of power filter has its Design procedures for fourth-ordered PPFs are illustrated in
own advantages and disadvantages, and PPFs are a simple [13], whereas passive LCL filters for APFs are designed in
yet effective solution. APFs are more popular because of [14]. Along with the optimal allocation of the PPFs, the
their dynamic ability to solve numerous power quality issues, feeder reconfiguration is also taken into account in [15]. An
but the cost of the equipment is very high compared to PPFs. adaptive version for the allocation, planning, and sizing of
The PPFs comprise passive elements like resistance (R), the PPFs is presented in [16]. Another Pareto-based method,
Inductance (L), and capacitance (C), which provide the namely the Bat algorithm, is applied for the design of PPFs
shortest impedance path for harmonics when properly in [17]. The comparative analysis of the optimal damped and
designed and configured. Based on the type and undamped PPFs is presented in [18].
configuration of the passive elements, PPFs are categorized
broadly as single-tuned, double-tuned, and C-type filters to In this paper, an improved design and performance
mitigate specific single-ordered, double-ordered, and higher- analysis for the single-tuned PPFs is presented using a robust
ordered harmonics, respectively. Out of these configurations, multi-objective mayfly optimization (RMOMO) method [19]
single-tuned PPFs are simple yet effective and less complex having the major objectives as minimized THD in current
and voltage, improved power factor, reduced losses of the

979-8-3503-9072-8/24/$31.00 ©2024 IEEE 1

Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.
filter, effective cost management, avoiding the resonance n

issues, and the reactive power compensation. ¦V

h 1
CP (h).il (h).cos Ih
PF (3)
The organization of the paper is as follows. Section II § n 2
n
2·
1/ 2

describes the system configuration, Section III illustrates the ¨ ¦ VCP (h) .¦ il (h) ¸
design and optimization procedure analysis, Section IV ©h 1 h 1 ¹
describes the simulation results and performance evaluation, where VCP is the voltage at the CPC, il is the line current,
and Section V presents the conclusion of the work. and ij is the angle between them at each order of harmonics
h. The total harmonic distortion (THD) in the line current
II. SYSTEM DESCRIPTION and voltage at the CPC can be calculated as,
Source
3ph Supply impedance Load
Line impedance
1/ 2
R Rs Ls isR ilR iLR Nonlinear §§ n 2· ·
¨ ¨ ¦ il (h) ¸ ¸
Rl Ll
B Y Rs Ls isY Rl Ll ilY CPC iLY
isB ilB iLB
%THD ¨¨ © h 2 ¹ ¸100
¸ (4)
Rs Ls Rl Ll
il1
¨ ¸
R5 R13
¨ ¸
L5 L13
Linear
© ¹
C5 C13
where il1 is the fundamental component of the line
PPF branch current, and vCP1 is the fundamental voltage at the CPC. The
Fig.1 System configuration
power losses in the PPF branches are considered mainly due
to the resistance inherited by the inductors or the resistances
Figure 1 represents the simple configuration of the that decide the quality factor (QF). Considering the
system where the distribution supply is represented as a Thevenin’s resistance of the PPF branch as RTH, the total
three-phase source with phases R, Y, and B with source and power losses caused by each harmonic are given by,
line impedance parameters as shown. The load of the system n
2
consists of both linear and nonlinear types. The single-tuned Ploss ¦ i ( h) l RTH (5)
PPF branch is connected individually at the common point of h 1

coupling (CPC), where the passive elements resistance (R), Minimizing the power losses will improve the efficiency
inductance (L), and capacitance (C), are designed to mitigate of the system. For the economical operation of the PPF
each specific ordered harmonic frequency individually. The equipment, the operational cost (OC) is considered in
orders can generally be of dominant frequencies, such as 5th, addition to the equipment investment cost (IC), which is
7th, 11th, and 13th. Single-tuned PPFs are simple to configure determined by the rating, value, and type of the passive
and cause less loss compared to other types, thereby elements. Thus, the total cost of the equipment (EC) is
minimizing the operational cost of the equipment. considered as,

III. DESIGN AND OPTIMIZATION ANALYSIS EC IC  OC Ÿ aR  bL  cC  dPloss  eQs (6)

A. Design Procedure where a, b, c, d, and e are the cost-weighting factors. The
For a series connected RLC branch of the PPF, the major constraints for the aforementioned parameters, as per
resonant frequency is the harmonic frequency that is to be the standards, to operate the PPF equipment effectively are
mitigated and can be expressed as, given below.
1 THD( R, L, C ) d 5% ; 0.9 d PF ( R, L, C ) d 1 (7)
XL X C 2S hf1 L (1)
2S hf1C
where f1 is the fundamental frequency of the system, and h is 20 d QF d 90; Qs min d Qs d Qs max (8)
the order of the harmonic to be mitigated. The value of C is The tuned frequency is to be 3% to 4% around the exact
to be designed first with respect to the amount of reactive harmonic frequency in order to avoid unwanted resonance
power to be compensated, which can be expressed as, with system parameters.
vs2 Qs
Qs 2S hf1C vs2 Ÿ C . (2) 1500
Impedance

Xc 2S hf1 vs2 1250

where Qs is the system reactive power requirement and vs

Impedance (ohms)

1000

is the system voltage. The L value can be obtained by 750

500
substituting (2) in (1). All L and C values of the branches of 250
the PPF are designed in a similar manner for different orders 0
of harmonic frequencies to be tuned. The impedance- 0 50 150 250 350 450 550 650 750 850 950

frequency characteristics of the PPF tuned for 5th, 7th, 11th, Frequency (Hz)
Phase
and 13th-ordered harmonics are shown in Figure 2. By 120

compensating for the reactive power requirement and 90

60
mitigating the harmonic content, the power factor of the 30
Phase (deg)

system can be improved, and it can be calculated as, 0

-30
-60
-90
-120
0 50 150 250 350 450 550 650 750 850 950

Frequency (Hz)

Fig.2 Impedance-Frequency curves of PPF

2
Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.
B. Optimization Analysis where R is the random value in a specified range. The initial
A robust multi-objective mayfly algorithm developed by velocities of the OS are zero. A detailed explanation of the
the authors of [19] is implemented for the optimum design of suggested method is presented in [19]. The flow diagram of
the single-tuned PPFs. Though numerous optimization the suggested method is shown in Figure 3.
algorithms exist in the literature, the RMOMO method is Start
chosen for its better convergence and speed during the
optimization process. The exploration is also improved by Define Objective Function
f(x), x=(x1, x2 ,. . . xN )T
Update solutions and velocities for both
male and female mayflies
the suggested method. This method is developed by the
inspiration of the mayflies whose flight behavior and mating Initialize the population and velocities of
male mayflies
Evaluate the solutions

process are considered and evaluated for the optimization xa (a = 1,2,. . . . N), ua
Give ranking for the mayflies
problem. In general, any optimization problem tries to Initialize the population and velocities of
minimize or maximize the value of an objective function, female mayflies Mate the bes t pairs according to ranking
& continue the process
yb (b = 1,2,. . . . M), ub
which can be expressed as follows:
Evaluate the solutions Evaluate the offspring functions (OS)
min or max f (x), x > x1 , x2 , x3 ,..., xM @ (9)
Separate the offspring of the mayflies
Find the solution for global best, Gbest
randomly
where M is the number of variables of a single objective
function f(x). For a multi-objective function, (9) can be Are the
stopping
No
Replace the solutions with best solutions

modified as, criteria met ?

Update Pbes t and Gbest
Yes

min or max F(x), x > f1 (x), f 2 (x), f3 (x)..., f N (x)@ (10) Do postprocess and vis ualize the results

where N is the number of objectives. The optimization Stop

algorithm tries to find the best set of nondominated solutions, Fig. 3 Flow diagram of the Mayfly algorithm
which is also known as the Pareto front, as there exists no
single best solution for a multi-objective optimization By using the RMOMO method, the objective functions
problem. According to the suggested method, the velocity u while designing the PPF are taken as follows. The overall
of a male mayfly a in the dimension d at a time step t is objective function of the suggested method is considered to
expressed as, find the set of minimum best solutions. As the power factor
should be maximum, it can be expressed as,
 E rp2  E rg2
t 1
uad t
g * uad  k1e t
Pbestad  xad  k 2e t
Gbestd  xad (11)
min f1 ( x) min ^1  PF ( R, L, C )` (16)
t 1 t 1
°­ x , if f ( x )  f ( Pbesta )
a a (12)
Pbesta ® The THD should be minimum in both line current and
°̄remain unchanged ; else
CPC voltage, and it is expressed as,
Gbest min^ f ( Pbest1 ), f ( Pbest2 ), f ( Pbest3 ),. . . ., f ( PbestM )` (13) min f 2 ( x) min THD( R, L,C ) (17)
where g is the fixed gravity coefficient in the range of The losses should be minimal, and it is expressed as,
(0,1], k1 and k2 are the attraction constants (>0), ȕ is the fixed
coefficient of visibility of the male mayfly a, rp is the min f 3 ( x) min Ploss ( R, L,C ) (18)
cartesian distance between the position xa and best position
Pbesta of the male mayfly a, and rg is the cartesian distance The total equipment cost should be minimal, and it is
between xa and best position Gbest of the male mayfly a . expressed as,
Similarly the velocity u of a female mayfly b in the min f 4 ( x) min EC ( R, L,C ) (19)
dimension d at a time step t is expressed as,
2 All the objective functions are to be defined by meeting
t
­ g * ubd  k 2 e  E rab xadt t
 ybd ; if male mayfly dominates female the constraints specified as above.
°
t 1 ° i.e. f ( yb ) ! f ( xa ) (13)
ubd ® IV. SIMULATION RESULTS AND ANALYSIS
t
° g * ubd  fl * r ; else
°
i.e. f ( yb ) d f ( xa ) The suggested method, as per the developed objective
¯
functions defined as aforementioned, is simulated using
where rab is the cartesian distance between the male and MATLAB/Simulink software, and the results are obtained as
female mayflies a and b, fl is the coefficient of random walk follows. The simulation parameters are listed in Table 1.
used when there is no female mayfly attracted to any male
mayfly, and yb is the position of the female mayfly b. Both TABLE I. SYSTEM PARAMETERS
male and female mayflies mate according to the objective Parameter Value
function defined to give the best possible solution for each
iteration. The first-best male and female are matched first; System supply voltage, vs (3ph, RMS, ph-ph) 415V
the second-best couple matches next, and this process Fundamental supply frequency , f1 50Hz
Source impedance, Rs , Ls Ÿ, 10mH
continues for no. of iterations. The crossover results with two
offspring (OS) are generated as follows. Line impedance, Rl, Ll 0.01 ŸP+
Load active and reactive power requirement PL,QL 1.5kW, 400VAR
OS1 R * a  (1  R ) * b (14) No. of single-tuned PPF branches 4
Nominal reactive power for each PPF branch 100VAR
OS 2 R * b  (1  R) * a (15) Order of harmonics to be mitigated 5th,7th, 11th &13th

3
Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.
 400 400 400
200 200 200

(V)
(V)
(V)

0
0 0

s
-200 -200
s

v
-200
v

-400 -400
-400

35 35 35
15 15

i (A)
i (A)
i (A)

0
-10 -10

s
s

s
-35 -35 -35
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29 0.3 0.35 0.36 0.37 0.38 0.39 0.4 0.41 0.42 0.43 0.44 0.45
Time (seconds) Time (seconds)
Time (seconds)

(a) (b) (c)

Signal
FFT window: 3 of 50 cycles of selected signal Signal
30
Signal
FFT window: 3 of 50 cycles of selected signal FFT window: 3 of 50 cycles of selected signal
10 40
s (A )

35
-10 10

i (A)

i (A)
-10 0
i

s
-30 -40
-35
0.02 0.03 0.04 0.05 0.06 0.07 0.25 0.26 0.27 0.28 0.29 0.3
0.4 0.41 0.42 0.43 0.44 0.45
Time (s) Time (s)
Time (s)
FFT analysis
FFT analysis FFT analysis
Fundamental is (50Hz) = 31.02A , THD= 22.57%
20 Fundamental i s (50Hz) = 30.93A , THD= 12.60% Fundamental i s (50Hz) = 30.97A , THD= 8.52%
20
20
15
Mag (% of Fundamental)

15
Mag (% of Fundamental)

Mag (% of Fundamental)
15
10
10
10

5
5 5

0
0 0
-1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic order Harmonic order Harmonic order

(d) (e) (f)

400
400
200
200 5
(V)

0
(V)

(A)
0 0
s

-200
v
s

-200
-5
v

5
-400

i
-400

40
40
20
20 5
i (A)

0
i (A)

(A)
-20 0
s

-20
s

-40
7
-40 -5
i

0.55 0.56 0.57 0.58 0.59 0.6 0.61 0.62 0.63 0.64 0.65
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85

Time (seconds) Time (seconds)

(g) (h)
Signal
Signal
FFT window: 3 of 50 cycles of selected signal
40 FFT window: 3 of 50 cycles of selected signal
20 20
5
(A)

0
(A)

0
(A)

-20
s

-20
0
s
i

-40
0.6 0.61 0.62 0.63 0.64 0.65 0.8 0.81 0.82 0.83 0.84 0.85
1 1

Time (s) Time (s)

-5
i

FFT analysis FFT analysis

Fundamental i (50Hz) = 30.96A , THD= 6.29% Fundamental i s (50Hz) = 31.02A , THD= 4.66%
s

20 20
5
(A)

0
Mag (% of Fundamental)

15
Mag (% of Fundamental)

15
1 3

10
10
-5
i

5
5
0.2 0.22 0.24 0.26 0.28 0.3
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 0 Time (seconds)
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic order
Harmonic order

(i) (j) (k)

Fig. 4 Waveforms of the supply voltage (vs), and line current (is); (a) when no filter is connected; (b) when only 5th ordered harmonics are mitigated; (c) 5th &
7th ordered harmonics are mitigated; (g) 5th, 7th, 11th ordered harmonics are mitigated; (h) 5th, 7th, 11th, 13th ordered harmonics are mitigated; (d),(e),(f),(i),(j)
corresponding harmonic spectrums of is.;(k) corresponding harmonic currents of 5th , 7th , 11th , 13th orders
Convergence
70
RMOMO
Generated solutions in the last iteration
1 F1
60 DE 0
GA Non-dominated solutions 10
Obtained PF
PSO Candidate solutions RMOMO
50 0.8 True PF
Objective Function

40 10
-100
0.6
Fi tness
F2

30
0.4
20 10
-200

10 0.2

0 0 10
-300
0 20 40 60 80 100 0 0.2 0.4 0.6 0.8 1 0 100 200 300 400 500
Iterations F1 Iterations
(a) (b) (c)

4
Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.
 THD,il Ploss
THD,VCP
0.049 0.179
0.049
0.178
0.047
0.047 0.177

OFG best
0.045

OFG best
0.176
0.045
OFG best

0.043 0.175
0.043
0.174
0.041
0.041 0.173

0.039 0.039 0.172

0 20 40 60 80 100 0 20 40 60 80 100 0 10 20 30 40 50
Iterations Iterations Iterations
(d) (e) (f)

EC PF
0.250
0.990
0.0117
0.248 0.989
0.0116
0.988
0.246
OFG best
OFG best

OFG best
0.987 0.0115
0.244
0.986
0.0114
0.242 0.985

0.240 0.984 0.0113

0.983
0.0112
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Iterations Iterations Iterations
(g) (h) (i)
Fig. 5 Results obtained by the RMOMO method; (a) convergence curves using different algorithms; (b) Pareto Front for two OFs; (c) Pareto Front for standard
fitness; (d) Gbest for THD in Vcp; (e) Gbest for THD in il ; (f) Gbest for Ploss; (g) Gbest for power factor; (h) Gbest for total equipment cost; (i) Gbest for
overall objective function

From the simulation results, as shown in Figure 4, it is The suggested optimization method is able to design the
observed that the source current, the supply current (is) passive element parameters of the PPF of each branch,
remains the same as the load current iL when no filter is whose values and optimized results are listed in Table 2.
applied till 0.1s, as shown in Figure 4(a). The harmonic Figure 5 shows the visualization of the optimization process
spectrum of is in this case is shown in Figure 4(d) where the for a selected number of iterations. It is observed that for a
THD in is is observed as 22.57%. The PPF branch to mitigate standard test function, the suggested method converges very
the 5th harmonic branch is connected at 0.1s, after which the fast within less no. of iterations when compared to some of
waveform changes and starts to attain a sinusoidal shape as the existing algorithms such as Genetic Algorithm (GA),
the 5th harmonic content is greatly reduced, whose Particle Swarm Optimization (PSO), and Differential
corresponding harmonic spectrum is as shown in Figure 4(e). Evolution (DE). The comparison is shown in Figure 5(a).
The THD in is is observed as 12.60%, with reduced 5th order The obtained solutions by the RMOMO method for two
harmonic content from no filter connection. The other objective functions and the nondominated solutions are
ordered harmonics remained unchanged. At 0.3s, another observed, as shown in Figure 5(b). Further, the fitness of the
tuned PPF branch is connected to mitigate 7th ordered standard test function is achieved by the RMOMO method,
harmonics, and the waveforms are observed as shown in as shown in Figure 5(c). Both these observations imply that
Figure (c), and the corresponding harmonic spectrum of is is the suggested method is able to solve the optimization
shown in Figure 4(f). in this case, the is became more problems efficiently, and the best values are obtained by
sinusoidal than earlier as the dominant harmonics are meeting the constraints within fewer iterations, which
mitigated with THD in is as 8.52%. As the THD is not yet reduces the overall optimization time. The obtained best
achieved below the standards (<5%), another PPF branch is solutions of the objective functions defined for minimum
connected in addition to the existing ones, such that the 11th %THD in Vcp and il are shown in Figures 5(d) and 5(e),
ordered harmonic contents are mitigated. The waveforms in respectively. The minimized cost function in units to reduce
this case, are as shown in Figure 4(g), and the corresponding the power losses and equipment cost are obtained as shown
harmonic spectrum of the is is shown in Figure 4(i). it is in Figures 5(f) and 5(g), respectively. The power factor is
observed that the is attains a more sinusoidal shape than the maximized to a value of 0.9903, as observed in Figure 5(h).
earlier cases with a great reduction of THD in is as 6.29%. The overall objective function of the suggested method
The 5th, 7th, and 11th ordered harmonic contents are reached its minimum value within 20 iterations, as shown in
significantly mitigated in this case. To bring the THD in is up Figure 5(i), which shows its robustness in achieving the
to the standards, further harmonics of order 13 are mitigated solutions despite the constraints.
by employing an additional PPF branch. In this case, the
more accurate sinusoidal shape of is is observed from Figure From the results, it is observed that the RMOMO is able
4(h) where the harmonic spectrum from Figure 4(i) confirms to design the PPF elements with the constraints specified and
the less THD in is as 4.66% having all tuned orders of the objective functions defined by giving global solutions
harmonics are significantly mitigated. The individual accurately with less convergence time, which proves its
harmonic currents are shown in Figure 4(k). efficacy.

5
Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.
 TABLE II. OPTIMIZED RESULTS [9] Igor D. Melo, José L.R. Pereira, Abilio M. Variz, Paulo F. Ribeiro,
“Allocation and sizing of single tuned passive filters in three-phase
distribution systems for power quality improvement,” in Electric
Values of Passive Elements by RMOMO method Power Systems Research, Vol. 180, 106128, ISSN 0378-7796, 2020.
for each ordered harmonics
[10] M. Bajaj, A. K. Singh, “Optimal design of passive power filter for
5th order 7th order 11th order 13th order enhancing the harmonic-constrained hosting capacity of renewable
DG systems,” in Computers & Electrical Engineering, Vol. 97,
Resistance 107646, ISSN 0045-7906, 2022.
Ÿ Ÿ Ÿ Ÿ
(R) [11] M. Mohammadi, A. M. Rozbahani, M. Montazeri, “Multi criteria
Inductance simultaneous planning of passive filters and distributed generation
24.36mH 12.42mH 5.03mH 3.60mH
(L) simultaneously in distribution system considering nonlinear loads
Capacitance with adaptive bacterial foraging optimization approach,”
16.63μF 16.63μF 16.63μF 16.63μF
(C) International Journal of Electrical Power & Energy Systems, Vol. 79,
Reactive Pages 253-262, ISSN 0142-0615, 2016.
1kVAr 1kVAr 1kVAr 1kVAr
Power (Q)
[12] Z. M. Ali, F. Q. Alenezi, S. S. Kandil, S. H.E. Abdel Aleem,
Objective functions “Practical considerations for reactive power sharing approaches
among multiple-arm passive filters in non-sinusoidal power systems,”
OF1 OF2 OF3 OF4 International Journal of Electrical Power & Energy Systems, Vol.
Obtained 103, Pages 660-675, ISSN 0142-0615, 2018.
solutions 0.04660 0.1741 0.2422 0.990 [13] N. M. Khattab, S. H.E. Abdel Aleem, Aboul'Fotouh El'Gharably,
(Gbest) Tarek A. Boghdady, Rania A. Turky, Ziad M. Ali, Mahmoud M.
Sayed, “A novel design of fourth-order harmonic passive filters for
total demand distortion minimization using crow spiral-based search
V. CONCLUSION algorithm,” Ain Shams Engineering Journal, Vol.13, Issue 3, 101632,
In this paper, a robust multi-objective mayfly ISSN 2090-4479, 2022.
optimization method is implemented to design the single- [14] M. Tavakoli Bina, E. Pashajavid, “An efficient procedure to design
passive LCL-filters for active power filters,” Electric Power Systems
tuned passive power filters to improve power quality. The Research, Vol. 79, Issue 4, Pages 606-614, ISSN 0378-7796, 2009.
optimization objective functions are defined mainly to reduce
[15] M. Sedighizadeh, R. Vali Doyran, A. Rezazadeh, “Optimal
the total harmonic distortions in voltage at the common point simultaneous allocation of passive filters and distributed generations
of coupling and line current, reduce the power losses, as well as feeder reconfiguration to improve power quality and
improve the power factor, and reduce the overall cost of the reliability in microgrids,” Journal of Cleaner Production, Vol. 265,
equipment. The suggested method is able to design the 121629, ISSN 0959-6526, 2020.
passive elements of the filter with respect to the constraints [16] M. Mohammadi, “Bacterial foraging optimization and adaptive
defined and give the optimum solution for all objective version for economically optimum sitting, sizing and harmonic tuning
orders setting of LC harmonic passive power filters in radial
functions defined. The suggested method proved to be robust distribution systems with linear and nonlinear loads,” Applied Soft
and fast convergent when compared to most of the traditional Computing, Vol. 29, Pages 345-356, ISSN 1568-4946, 2015.
algorithms. [17] N. Yang, M. Le, “Optimal design of passive power filters based on
multi-objective bat algorithm and pareto front,” Applied Soft
REFERENCES Computing, Vol. 35, Pages 257-266, ISSN 1568-4946, 2015.
[1] J. C. Das, “ Power system Harmonics and passive filter designs,”
Wiley: NJ, 2015. [18] N. Abdul Kahar, A. F. Zobaa, Rania A. Turky, Ahmed M. Zobaa,
Shady H.E. Abdel Aleem, Bazilah Ismail, “Comparative analysis of
[2] Z. Cui, C. Li, W. Dai, L. Zhang and Y. Wu, "A Hierarchical
optimal damped and undamped passive filters using MIDACO-
Teaching-Learning-Based Optimization Algorithm for Optimal
solver,” Ain Shams Engineering Journal, Vol.14, Issue 8, 102056,
Design of Hybrid Active Power Filter," in IEEE Access, vol. 8, pp.
2023.
143530-143544, 2020.
[19] K. Zervoudakis, S.Tsafarakis, “A mayfly optimization algorithm,”
[3] Y. Wu, R. Chen, C. Li, L. Zhang and W. Dai, "An Adaptive Sine- Computers & Industrial Engineering, Volume 145, 106559, ISSN
Cosine Moth-Flame Optimization Algorithm for Parameter 0360-8352, 2020.
Identification of Hybrid Active Power Filters in Power Systems,"
in IEEE Access, vol. 8, pp. 156378-156393, 2020.
[4] J. Huang, C. Li, Z. Cui, L. Zhang and W. Dai, "An Improved
Grasshopper Optimization Algorithm for Optimizing Hybrid Active
Power Filters’ Parameters," in IEEE Access, vol. 8, pp. 137004-
137018, 2020.
[5] L. Zhang, C. Li, Y. Wu, J. Huang and Z. Cui, "An Improved Salp
Swarm Algorithm With Spiral Flight Search for Optimizing Hybrid
Active Power Filters’ Parameters," in IEEE Access, vol. 8, pp.
154816-154832, 2020.
[6] S. H. E. Abdel Aleem, A. F. Zobaa, M. E. Balci and S. M. Ismael,
"Harmonic Overloading Minimization of Frequency-Dependent
Components in Harmonics Polluted Distribution Systems Using
Harris Hawks Optimization Algorithm," in IEEE Access, vol. 7, pp.
100824-100837, 2019.
[7] A. F. Zobaa, "Mixed-Integer Distributed Ant Colony Multi-Objective
Optimization of Single-Tuned Passive Harmonic Filter Parameters,"
in IEEE Access, vol. 7, pp. 44862-44870, 2019.
[8] M. Bajaj, N. K. Sharma, M. Pushkarna, H. Malik, M. A. Alotaibi and
A. Almutairi, "Optimal Design of Passive Power Filter Using Multi-
Objective Pareto-Based Firefly Algorithm and Analysis Under
Background and Load-Side’s Nonlinearity," in IEEE Access, vol. 9,
pp. 22724-22744, 2021.

6
Authorized licensed use limited to: Yonsei Univ. Downloaded on December 19,2024 at 12:34:06 UTC from IEEE Xplore. Restrictions apply.