applied

sciences
Article
Differential Evolution Applied to a Multilevel Inverter—A
Case Study
Oscar Sánchez Vargas, Susana Estefany De León Aldaco * , Jesús Aguayo Alquicira , Luis Gerardo Vela Valdés
and Jesús Darío Mina Antonio

Tecnológico Nacional de México/Centro Nacional de Investigación y Desarrollo Tecnológico (CENIDET),

Cuernavaca 62490, Mexico
* Correspondence: [Link]@[Link]

Abstract: This work aims to obtain the switching angles to reduce the total harmonic distortion
(THD) and compare the different sets of angles obtained by modifying the modular index (Mi)
in a seven-level multilevel inverter using the differential evolution (DE) algorithm. A seven-level
cascaded multilevel inverter with RL load was simulated in MATLAB Simulink. The Mi was varied
(0.6 to 1.27) to check which set of switching angles is optimal to reduce the THD. These sets of angles
are then experimentally validated on a test bench. A brief discussion of the results is presented,
comparing the minimum THD found without considering Mi, versus the minimum THD obtained if
Mi is less than 1.

Keywords: CHBMLI; differential evolution; metaheuristic methods; multi-level inverter; THD

reduction

Citation: Sánchez Vargas, O.; 1. Introduction

De León Aldaco, S.E.; Aguayo Metaheuristic algorithms are optimization algorithms attempt to improve the search
Alquicira, J.; Vela Valdés, L.G.; Mina
for solutions in the search space iteratively. The algorithms may be inspired by biological
Antonio, J.D. Differential Evolution
phenomena or by the social behavior of species. Metaheuristic algorithms are usually
Applied to a Multilevel Inverter—A
problem-independent. Due to their specific advantages over classical methods, metaheuris-
Case Study. Appl. Sci. 2022, 12, 9910.
tics are used in the solution of different problems in the fields of engineering [1].
[Link]
Among the metaheuristic methods that have been used for the solution of these
Academic Editor: Graziella problems are differential evolution, particle optimization (PSO), genetic algorithm (GA),
Scandurra firefly algorithm (FA), and others. However, the differential evolution algorithm has
Received: 8 September 2022
distinguished itself for its efficiency and belongs to the evolutionary algorithms. Among
Accepted: 29 September 2022
the outstanding features of the differential evolution algorithm are its efficient global search
Published: 1 October 2022
capability, versatility to the characteristics of the problem, and not requiring a suitable
initial solution [2].
Publisher’s Note: MDPI stays neutral
Differential evolution has been used in various engineering fields, such as acoustics,
with regard to jurisdictional claims in
aerodynamics, aerospace, agriculture, automotive, biology, chemistry, defense, and others.
published maps and institutional affil-
The differential evolution algorithm has been widely used in electricals and electronics, in
iations.
various communication systems (multi-input and multi-output systems), power allocation,
system modeling, controller design for complex dynamic systems, radio network design,
ultra-wideband radio systems, and power converters [3].
Copyright: © 2022 by the authors.
Multilevel inverters (MLIs) can create a single output waveform by synthesizing
Licensee MDPI, Basel, Switzerland. multiple voltage levels, producing a stepped waveform approximating a sinusoidal signal.
This article is an open access article As the number of sources at the input to the inverter increases, so does the number of levels
distributed under the terms and of the output waveform [4,5]. Due to the stepped nature of the multilevel inverter output
conditions of the Creative Commons waveform, multiple sets of switching angles can be found for one modulation index (Mi)
Attribution (CC BY) license (https:// value [6].
[Link]/licenses/by/ For applications related to power converters, the differential evolution algorithm has been
4.0/). used to minimize THD in multilevel inverters by calculating the optimal switching angles.

Appl. Sci. 2022, 12, 9910. [Link] [Link]

Appl. Sci. 2022, 12, 9910 2 of 15

For applications related to power converters, the differential evolution algorithm has
Appl. Sci. 2022, 12, 9910 been used to minimize THD in multilevel inverters by calculating the optimal switching 2 of 15
angles.
The solution of the transcendental nonlinear equations is one of the main problems
in calculating
The solutionthe optimal switching angles
of the transcendental to minimize
nonlinear [Link],
equations one of thedifferent optimi-
main problems
zation methods have been used in the literature to eliminate harmonics
in calculating the optimal switching angles to minimize THD. Therefore, different opti- by solving the
nonlinear transcendental equations that generate the Fourier series.
mization methods have been used in the literature to eliminate harmonics by solving the
Several
nonlinear papers mention
transcendental performing
equations a sweep of
that generate thethe Mi and
Fourier obtaining results of a min-
series.
imumSeveral
THD using
papers a Mi > 1 [6–8],
mention which can
performing produce
a sweep of an
theovermodulation,
Mi and obtaining andresults
they only
of a
considered
minimum THD using the aminimum
using Mi > 1 [6–8],%THD whichfound. In otheranpapers
can produce [9–12], “coincidentally”,
overmodulation, and they only
the minimum
considered %THD
using found had
the minimum a Mi>found.
%THD 0.6 and In<1. Reviewing
other the angles
papers [9–12], found in these
“coincidentally”, the
papers that coincide in the number of levels (7), they obtain
minimum %THD found had a Mi > 0.6 and <1. Reviewing the angles found in these a set of switching angles dif-
papers
ferent from each
that coincide other.
in the number Additionally,
of levels (7),other
theyarticles
obtain ado setnot mention which
of switching anglesMi they used.
different from
Moreover,
each other. they have different
Additionally, other levels,
articlesloads,
do notand phases
mention implemented
which in theMoreover,
Mi they used. MLIs. they
haveThis paperlevels,
different focuses on the
loads, anddifferential evolution (DE)
phases implemented in thealgorithm
MLIs. application to mini-
mize This
the THD
papervalue
focusesin the output
on the voltageevolution
differential waveform(DE) of aalgorithm
cascaded multilevel
applicationinverter and
to minimize
to
theanalyze
THD valuethe obtained
in the outputsets ofvoltage
switching angles calculated
waveform of a cascadedto achieve a required
multilevel modular
inverter and to
index
analyze according to thesets
the obtained inverter application.
of switching angles calculated to achieve a required modular
index according to the inverter application.
2. Materials and Methods
2. Materials and Methods
2.1. Differential Evolution Algorithm
2.1. Differential Evolution Algorithm
The differential evolution (DE) algorithm is a relatively new metaheuristic method.
The DEThealgorithm
differential is evolution (DE) algorithm
a population-based is a relatively
algorithm that usesnew metaheuristic
genetic operators method.
such as
The DE algorithm is a population-based algorithm that uses genetic operators such as
crossover, mutation, and selection.
crossover, mutation, and selection.
It has become one of the most widely used evolutionary algorithms. It is easily un-
It has become one of the most widely used evolutionary algorithms. It is easily
derstood and easy to implement.
understood and easy to implement.
Kenneth and Storn developed the differential evolution algorithm as a reliable, ver-
Kenneth and Storn developed the differential evolution algorithm as a reliable, ver-
satile, easy-to-implement function optimizer. The first publication on DE appeared in a
satile, easy-to-implement function optimizer. The first publication on DE appeared in a
technical report in 1995 [13].
technical report in 1995 [13].
This metaheuristic method has three principal operations: mutation, crossover, and
This metaheuristic method has three principal operations: mutation, crossover, and
selection. In the mutation stage, the population vectors are randomly perturbed. Popula-
selection. In the mutation stage, the population vectors are randomly perturbed. Population
tion diversity is controlled by the crossover process [14]. Meanwhile, the selection process
diversity is controlled by the crossover process [14]. Meanwhile, the selection process
chooses the vectors containing the fittest solution [15].
chooses the vectors containing the fittest solution [15].
Some
Some advantages
advantages of of the
the DE
DE method
method are are its
its ease
ease of
of application,
application, fast
fast convergence,
convergence, andand
tendency
tendency to reach the global optimal solution without remaining in localsolutions
to reach the global optimal solution without remaining in local solutions[16,17].
[16,17].
A
A diagram
diagram of of the
the stages
stages of of DE
DE isis shown
shown in in Figure
Figure 1. 1.

Diagram of
Figure 1. Diagram
Figure of the
the stages
stages of differential evolution.

The following is a brief description of each stage outlined in the diagram in Figure 1:
1. Initialization (Initial population): A population of vectors is randomly generated
(individuals) [18–20].
Appl. Sci. 2022, 12, 9910 3 of 15

n o
Let SG = X jG : j = 1, 2, . . . , NP (1)

be the population of any generation of G, NP denotes the population size. D is the length
of the search space. n o
X jG = G
x1,j G
, x2,j G
, . . . , x D,j · X jG (2)

is generated using a uniformly distributed random number, such as: “rand (0,1)”. Hence,
we obtain:
X jG = X Low + XU pp − X Low ∗ rand (0, 1 )


(3)
where:
XLow = Lower bound.
XUpp = Upper bound.
SG = search space.
2. Mutation: In this process, three vectors are randomly selected, and the first two vectors
are subtracted from each other (this is to define a search direction). The difference is
multiplied by the scale factor or “F”, which can vary between zero and one [12]. To
the resulting vector, the third vector is added [21], as in the following equation:
 
VjG = Xr3
G
+ F ∗ Xr1 G
− Xr2G
(4)

where Vj G is the mutant vector.

For each vector in the population, a mutated vector is created.
3. Crossover: A new vector called the test vector (child vector) is generated, using a
crossover factor Cr with values between 0 and 1 and defining the degree of similarity
of the test vector to the mutant or parent vector. If Cr is close to 1, the test vector will
be quite similar to the mutant vector; if Cr is close to 0, it will be similar to the parent
vector [19,20].
The test vector Uj G , the target vector Xj G , and the mutant vector Vj G are denoted as:
n o
UjG = u1,jG G
, u2,j , . . . , uG
D,j

n o
X jG = G
x1,j G
, x2,j G
, . . . , x D,j
n o
VjG = v1,j
G G
, v2,j , . . . , vG
D,j

so Uj G is generated as:
(
G I f rand ≤ Cr
vi,j
G j
ui,j = G Otherwise (5)
xi,j

where i ∈ {1, 2, . . . , D}.

4. Selection: The target vector is compared with the test vector. The vector with the
highest fitness remains in the next generation, while the vector with the lowest fitness
is eliminated [3,20]. This operation is performed as:
(    
G +1 UjG I f f UjG ≤ f UjG
Xj = (6)
X jG Otherwise

Scale factor (F) influences the search’s orientation and convergence speed. If the F
factor takes values between 0.1 and 0.4, it allows a better local search in the space. If F
takes a value between 0.4 and 0.95, the algorithm performs a better global search. Some
authors recommend using 0.9, as it maintains a balance between speed and probability of
convergence [15,22,23].
Appl. Sci. 2022, 12, 9910 4 of 15

Several authors recommend using a crossover rate (Cr) with a value of 0.9 so that the
test vector will be 90% similar to the mutant vector and thus include a wide variety of
individuals [24].

2.2. Optimization Problem Statement

Optimization is obtaining the best possible solution to several designs and
engineering problems.
An optimization problem contains three important aspects:
1. Objective function: the property to be optimized, which can be expressed as a linear
or non-linear function.
2. Decision variable: an unknown element of an optimization problem.
3. Constraints: restrictions that must be satisfied to produce an acceptable result.
For a stated objective function that is defined on the set of all decision variables, a de-
cision must be found such that the objective is optimized (minimized or maximized) [1,25].
An optimization problem may be as follows:
min f ( x ) , subject to x ∈ Ω (7)

where “f ” is a real-valued function called the objective function.

The vector “x” is a vector of “n” independent variables x = [x1 , x2 . . . xn ] where the
variables x1 , x2 . . . xn , often referred to as decision variables, are selected by the user or
designer. The set Ω is a subset of R called the constraint set or feasible set.
The above problem is a decision problem. It involves finding the “best” vector x in
compliance with the constraints, or nay, all possible vectors within Ω. The “best” vector
is the one that has the smallest value of the objective function. This vector is called the
“minimizer”. It is possible to have many minimizers.
Most engineering problems require the designer to optimize some conflicting objec-
tives. Objectives conflict if one objective’s improvement leads to another’s deterioration.
The simplest way to optimize is to find a decision variable that satisfies the given constraints
and optimizes a function whose components are objective functions [25,26]. Multi-objective
problems may not have a single optimal solution.
In this paper, the optimization problem is the reduction of THD in a cascade H-
bridge multilevel inverter (CHBMLI) and finding a desired value of Mi according to the
application. Starting from the THD, the stepped waveform of the output voltage is analyzed
using the Fourier series and is expressed in the following equation, considering that it has
quarter-wave symmetry:
∞
Vout (θ ) = ∑n=1,3,5 bn sin(nθ ) (8)
where n = 1, 3, 5 are odd harmonics, and bn is given by:
2N −1 4Vdc
bn = ∑n=1,3,5 nπ
(V1 cos(nα1 ) + V2 cos(nα2 ) + · · · Vm cos(na N )) (9)

where:
N = number of switching angles per quarter cycle.
n = 1, 3, 5 . . . 2N −1 (odd harmonics).
m = number of DC sources.
α = switching angle.
The quality of the output voltage waveform is determined by how many harmonics it
contains, so Equation (4) shows the THD formula following the IEEE 519 standard to be set
as an objective function: q
∑50 2
n=1,3,5,7... Vn
THD = ∗ 100 (10)
V1
where:
4Vdc 4V 4Vdc
V= cos(nα1 ) + dc cos(nα2 ) + . . . cos(nαn ),
nπ nπ nπ
 where:

𝑉= cos(𝑛𝛼 ) + cos(𝑛𝛼 )+… cos(𝑛𝛼 ),

Appl. Sci. 2022, 12, 9910 5 of 15
The angles must satisfy the following constraints:
𝜋
0≤𝛼 ≤𝛼 …≤𝛼 ≤ (11)
The angles must satisfy the following constraints: 2
where n = number of switching angles per quarter cycleπand α1 = switching angle 1, α2 =
switching angle 2, and αn = switching 0 ≤ α1angle
≤ α2 n. . (The
. ≤ αswitching
n ≤ angles are the decision var- (11)
2
iables).
where n = Mi
When number ofsaid
> 1, it is switching angles
that there per quarter cycle
is over-modulation, and results
which α1 = switching angleof1,
in the absence
α = switching angle 2, and α = switching angle n (The switching
pulses when comparing the modulating signal with the carrier signal.
2 n angles are the decision variables).
When Mi >Mi,
Regarding 1, itthe
is modulation
said that there is over-modulation,
ratio is equal to V1 overwhich
n. results in the absence of
pulses when comparing the modulating signal with the carrier signal.
Recalling that the case study is a seven-level CHBMLI, Figure 2 shows a waveform
Regarding
with seven levelsMi,
and the modulation
three switching ratio is equal
angles to V1
(n = 3); theover n. remains:
equation
Recalling that the case study is a seven-level CHBMLI, Figure 2 shows a waveform
with seven levels and three 4
𝑀𝑖switching
= angles
(𝑐𝑜𝑠(∝ ) +(n𝑐𝑜𝑠(∝
= 3); the
) +equation
𝑐𝑜𝑠 (∝ ))remains: (12)
3𝜋
4
Mi = (cos(∝1 ) + cos(∝2 ) + cos(∝3 )) (12)
3π

Figure
Figure 2.
2. CHBMLI
CHBMLIwaveform
waveformwith
with seven
seven levels
levels and
and three
three switching
switching angles.
angles.

2.3. Structure
2.3. Structureof
of the
the Seven-Level
Seven-Level Cascaded
Cascaded Multilevel
Multilevel Inverter:
Inverter: Case
Case Study
Study
Reference [27]
Reference [27] provides
providesa asystematic review
systematic reviewof publications usingusing
of publications metaheuristic meth-
metaheuristic
methods to minimize THD in multilevel inverters. The publication highlights thatmost
ods to minimize THD in multilevel inverters. The publication highlights that the the
used used
most topology in combination
topology with with
in combination the differential evolution
the differential algorithm
evolution is theissingle-phase
algorithm the single-
cascaded
phase multilevel
cascaded inverter
multilevel with seven
inverter [Link].
with seven
The most
The most outstanding
outstanding features
features of
of CHBMLI
CHBMLI are:
are:
•• The generation
The generationofofitsitsswitching
switching signals
signals is simple,
is simple, allowing
allowing the of
the use use ofPWM
any any PWM
tech-
technique
nique based
based on multi-carrier
on multi-carrier or vector
or vector spaces
spaces [28,29].
[28,29].
•• Switching states
Switching states can
can be be changed
changed toto compensate
compensate for
for faults.
faults.
•• Input capacitors have no voltage balance problems
Input capacitors have no voltage balance problems [30]. [30].
Considering the advantages, the topology selected as a case study for this paper is
the seven-level single-phase CHBMLI with an RL load connected to the output, as shown
in Figure 3. MATLAB Simulink software was used to perform the simulation of the case
study. The differential evolution method is used to minimize THD.
 Considering the advantages, the topology selected as a case study for
the seven-level single-phase CHBMLI with an RL load connected to the outp
Appl. Sci. 2022, 12, 9910 in Figure 3. MATLAB Simulink software was used to perform the 6 ofsimulati
15
study. The differential evolution method is used to minimize THD.

Figure 3. CHBMLI inverter with RL load.

Figure 3. CHBMLI inverter with RL load.
Table 1 shows the parameters used for the design and simulation of the cascaded
multilevel
Tableinverter:
1 shows the parameters used for the design and simulation of
multilevel inverter:
Table 1. Most important parameters for the simulation of CHBMLI in MATLAB.

Parameters Specifications
Table 1. Most important parameters for the simulation83.33
Voltage sources (Vdc 1,2,3)
of CHBMLI
V
in MATLAB.
R Load 100 Ω
Parameters
L Load 100 mH Specifications
Power
Voltage sources (Vdc 1,2,3) 625 W 83.33 V
Peak Voltage (Vout ) 250 V
R Load
Frequency 60 Hz 100 Ω
L Load 100 mH
3. Simulation Results Power 625 W
Peak Voltage
MATLAB software was used(Vto out)
program the DE algorithm. 250 V
For this case study, three
Frequencyswitching angles (∝1, ∝2, and ∝3) were searched for
60 the
Hz
seven-level CHBMLI.
After obtaining the different sets of switching angles, the modulation index sweep
3.
wasSimulation
performed. Results
Figure 4 shownn the simulation diagram of the seven-level single-phase CHBMLI,
MATLAB
was performed software
in MATLAB was used
Simulink to using
software program the DE inalgorithm.
the parameters Table 1.
For this
Figure case
5 shows thestudy, three switching
output voltage angles
waveform realized (∝1, ∝2,
in MATLAB and ∝3) were sea
Simulink.
As shown in
seven-level [Link] 5, the output peak voltage obtained is 250 V.
The initialization parameters of the differential evolution algorithm reported in [31]
After obtaining the different sets of switching angles, the modulation
were taken as a starting point. The parameters used in the case study of this article are
was
shown performed.
in Table 2.
Figure 4 shownn the simulation diagram of the seven-level single-ph
was performed in MATLAB Simulink software using the parameters in Tab
Appl. Sci. 2022, 12, 9910
Appl. Sci. 2022, 12, 9910 7 of 15

Appl. Sci. 2022, 12, 9910 7 of 15

Figure 4. The schematic diagram for CHBMLI Simulink simulation.

FigureFigure
Figure The 5schematic
4. The
4. shows the
schematic output
diagram forvoltage
diagram CHBMLI waveform
Simulink
for CHBMLI realized simulation.
in MATLAB Simulink.
simulation.
Simulink

Figure 5 shows the output voltage waveform realized in MATLAB Simulink.

Figure 5.
Figure 5. Output
Output peak
peak voltage
voltage in
in the
the CHBMLI
CHBMLI Simulink
Simulink simulation.
simulation.

As shown in Figure 5, the output peak voltage obtained is 250 V.

The initialization parameters of the differential evolution algorithm reported in [31]
were taken as a starting point. The parameters used in the case study of this article are
shown in Table 2.
Figure 5. Output peak voltage in the CHBMLI Simulink simulation.
 Appl. Sci. 2022, 12, 9910 8 of 15

Appl. Sci. 2022, 12, 9910 8 of 15

Table 2. Parameter values for the differential evolution algorithm.
Appl. Sci. 2022, 12, 9910 8 of 15
Parameter Value
Table 2. Parameter values for(NP)
Population the differential evolution algorithm. 100
Table 2. Parameter
Generation (G) values for the differential evolution algorithm.
100
Parameter
Scaling Factor (F)Parameter Value
0.3 Value
Populationrate
Crossover (NP)(Cr)
Population (NP) 100
0.9 100
Generation (G)Generation (G) 100 100
Scaling Factor Scaling
(F) 0.3
Figure 6 shows the step-by-step
Factorprocedure
(F) of the methodology that0.3
was carried out
Crossover rateCrossover
(Cr) rate (Cr) 0.9 0.9
to obtain the different sets of angles, the Mi, and the THD percentage.

Figure 6 showsFigure 6 shows the step-by-step

the step-by-step procedure ofprocedure of the methodology
the methodology that wasout
that was carried carried
to out
• Set limits [0° to 90°] The display shows the
obtain •theSet to obtain
different setsthe different sets of angles, the Mi, and the THD percentage.
parameters of of angles, the Mi, and the THD percentage. switching angles, the
Run the
individuals: G=100, best THD found and the
Start algorithm in
NP=50,F= 0.3• andSet limits [0° to 90°] Mi [Link] display shows the
• Set parameters of
Matlab
Cr=0.9 Run the switching angles, the
• Set Mi=0.66 individuals: G=100, best THD found and the
Start algorithm in
NP=50,F= 0.3 and Mi reached.
Matlab
Cr=0.9
• Set Mi=0.66
No No

No No
The best THD,
Mi>=1.27 Increases Mi 0.01 Yes Iterations >20 switching angles
The best THD,
and Mi are noted.
Mi>=1.27 Increases Mi 0.01 Yes Iterations >20 switching angles
and Mi are noted.
Yes
Yes
A graph of the magnitude of the
harmonics (1 to A50),graph
the %THD and
of the magnitude of the
The angles obtained are the fundamental signal is displayed
harmonics (1 to 50),onthe %THD and
The data obtained are
used in the simulationThe the screen.
of angles obtained are (Using
thethe Powerguisignal
fundamental FFTis displayed on End
used in the simulation of the screen. (Using the Powergui analysed
FFT andThe
plotted.
data obtained are
End
CHBMLI in Simulink. Analysis Tool). analysed and plotted.
CHBMLI in Simulink. Analysis Tool).

Figure 6. Flow chart

Figureof the methodology used.
Figure 6. Flow chart [Link]
Flow chart of the methodology
methodology used. used.

Theresults
The resultsof
ofThe
theseven-level
the seven-level
results CHBMLIpreviously
of the CHBMLI
seven-level previously simulated
CHBMLI previously
simulated inSimulink
Simulink
simulated
in weretested
in Simulink
were tested
were tested
and plotted to observe
and plotted its
tobehavior.
observe itsThe following
behavior. The figure
followingshows
figurethe results
shows
and plotted to observe its behavior. The following figure shows the results of the angles the of the
results angles
of the angles
obtained by obtained
the by the implementation
implementation of this of this algorithm.
algorithm.
obtained by the implementation of this algorithm.
Figure77shows
showsFigure 7graph
showsof the graph of the sets
different sets ofobtained
angles obtained (Angle 1 circles,
Figure thethegraph of the the different
different of angles
sets of angles obtained (Angle(Angle 1 circles,
1 circles, Angle
Angle 2 squares, and Angle 3 triangles) that coincide with some Mi.
2Angle 2 squares,
squares, and3Angle
and Angle 3 triangles)
triangles) that coincide
that coincide with some withMi.
some Mi.

Figure 7. Angles vs. Mi in CHBMLI Simulink simulation.

Appl. Sci. 2022, 12, 9910 9 of 15

Appl. Sci. 2022, 12, 9910 Figure 7. Angles vs. Mi in CHBMLI Simulink simulation. 9 of 15

The set of angles depends on Mi; the smaller the Mi, the more sets are available. Two
sets of
Theangles
set ofshowing the beston
angles depends angles (with
Mi; the lowest
smaller theTHD) found
Mi, the morefor a Mi
sets areof 0.97 (marked
available. Two
with black color) and 1.05 (marked with red color) have been marked in Figure
sets of angles showing the best angles (with lowest THD) found for a Mi of 0.97 (marked 7 to gen-
erateblack
with the following
color) andTable.
1.05 (marked with red color) have been marked in Figure 7 to generate
the following Table.
InInFigure
Figure8,8,only
onlythe
thelowest
lowestTHD
THDobtained
obtainedfrom
fromthe
theset
setof
ofangles
angles corresponding
corresponding to
to
each modular index was used, considering that in Figure 7, there are four combinations
each modular index was used, considering that in Figure 7, there are four combinations of
of angles
angles forfor
oneone
[Link].

Figure 8.
Figure 8. THD
THD vs.
vs. Mi
Mi in
in CHBMLI
CHBMLI simulink
simulink simulation.
simulation.

As can
As can be
be seen
seen in
in Figure
Figure 8,
8, the
the Mi
Mi that reached
reached the
the minimum
minimum percentage
percentage of of THD
THD was
was
1.05 (10.46%).
1.05 (10.46%).
Several articles
Several articles do
do not
not consider
consider thethe value
value ofof Mi;
Mi; there
there isis aa probability
probability ofof not
not finding
finding
the
the optimal
optimal angles
angles to
to reduce
reduce the
the THD or generate an overmodulation. For this case study,
the
the values
values with
with aa Mi
Mi << 1 were a Mi of of 0.97
0.97 and
and aa THD
THD of of 12.98%.
12.98%. It is shown in Figure 8.
marked
marked with
with red
red dots.
dots.
The
The following
following figures
figures compare
compare the the harmonic
harmonic spectrum
spectrum generated
generated by by the
the MATLAB
MATLAB
Simulink
Simulink Toolbox, from the Mi < 1 and regardless of the Mi, showing the percentage of
Toolbox, from the Mi < 1 and regardless of the Mi, showing the percentage of
THD
THD corresponding
correspondingto tothese.
these.
By
By not
notconsidering
consideringthe theMi,
Mi,there is is
there a high probability
a high probabilityof finding a seta of
of finding setangles that that
of angles can
generate an overmodulation
can generate an overmodulationwith awith
verya low
veryTHDlow but
THD is inefficient for the for
but is inefficient desired load.
the desired
Appl. Sci. 2022, 12, 9910
[Link] the above, looking at the values with a Mi < 1, the Mi with10a of low THD
15
percentage was a Mi of 0.97, with a THD of 12.98%, as shown in Figure
Considering the above, looking at the values with a Mi < 1, the Mi with a low THD 9b). Almost a 3%
difference from the THD percentage was found without considering the
percentage was a Mi of 0.97, with a THD of 12.98%, as shown in Figure 9b). Almost a 3% Mi (10.46%).
difference from the THD percentage was found without considering the Mi (10.46%).

(a) (b)
Figure 9. Harmonic spectrum generated with the MATLAB Simulink Toolbox: (a) harmonic spec-
Figure 9. Harmonic spectrum generated with the MATLAB Simulink Toolbox: (a) harmonic spectrum
trum generated with Mi 0.97; (b) harmonic spectrum generated with Mi 1.05 (minimum %THD
generated
found). with Mi 0.97; (b) harmonic spectrum generated with Mi 1.05 (minimum %THD found).

4. Experimental Results
A cascaded modular multilevel inverter was built to obtain the behavior of the
proposed method and THD, resulting in a stepped output voltage of seven levels. A three-
H-bridge configuration was used to generate seven output voltage levels. Figure 10 shows
 (a) (b)
Figure 9. Harmonic spectrum generated with the MATLAB Simulink Toolbox: (a) harmonic spec-
Appl. Sci. 2022, 12, 9910 trum generated with Mi 0.97; (b) harmonic spectrum generated with Mi 1.05 (minimum10 of 15
%THD
found).

[Link]
ExperimentalResults
Results
AAcascaded
cascadedmodular
modularmultilevel
multilevel inverter
inverter was
was built
built to obtain
to obtain the behavior
the behavior of theofpro-
the
proposed method and THD, resulting in a stepped output voltage of seven
posed method and THD, resulting in a stepped output voltage of seven levels. A three-H- levels. A three-
H-bridge
bridge configuration
configuration waswas used
used to generate
to generate seven
seven output
output voltage
voltage levels.
levels. Figure
Figure 10 10 shows
shows a
a CHBMLI
CHBMLI cell,cell,
usingusing an IRAMS10UP60A
an IRAMS10UP60A module module with
with its its respective
respective driver circuits,
driver circuits, imple-
implemented
mented with
with a 120 a 120
V DC V DC
power power
supply. Thesupply. The
switching switching
angles obtained angles
wereobtained
implemented were
inimplemented
the ALTERAinFPGA,the ALTERA FPGA, tables
using look-up using look-up tablesapplication.
for an offline for an offline application.
Figure 11 shows Figure
the
11 shows theplatform.
experimental experimental platform.

Appl. Sci. 2022, 12, 9910 11 of 15

Figure10.
Figure [Link]
ExperimentalH-Bridge
H-Bridgecircuit
circuitusing
usingIRAMS10UP60A
IRAMS10UP60Amodule
modulewith
withits
itsdriver
drivercircuit.
circuit.

Figure11.
Figure [Link]
Experimentalplatform,
platform,multilevel
multileveltopology.
topology.

Figure12
Figure 12shows
showsthe
theimplemented
implemented switching
switching signals
signals obtained
obtained experimentally
experimentally usingusing
the
set
theofset
angles in Table
of angles 3. It [Link]
in Table in the in
It is observed figure that the
the figure lines
that theare composed
lines of 3 colors;
are composed of 3
colors; the blue color indicates that it is off, the green color indicates that it is on, and the
yellow color indicates a signal being in transition.

Table 3. Best sets of angles found and their Mi.

 Figure 12 shows the implemented switching signals obtained experimentally using
the set of angles in Table 3. It is observed in the figure that the lines are composed of 3
colors; the blue color indicates that it is off, the green color indicates that it is on, and the
yellow color indicates a signal being in transition.
Appl. Sci. 2022, 12, 9910 11 of 15
Table 3. Best sets of angles found and their Mi.

Mi ∝1 ∝2 ∝3
0.97 indicates that it is9.80°
the blue color 28.63° that it is on, and
off, the green color indicates 64.2°
the yellow
1.05 8.69° 27.89° 49.81°
color indicates a signal being in transition.

Figure 12. Experimental H-Bridge output signals.

Figure 12. Experimental H-Bridge output signals.

Table 3. Best sets of angles found and their Mi.

Mi ∝1 ∝2 ∝3
Appl. Sci. 2022, 12, 9910 0.97 9.80◦ 28.63◦ 64.2◦ 12 of 15
1.05 8.69◦ 27.89◦ 49.81◦

Figure 13
Figure 13shows
showsthe
theoutput
outputvoltage
voltagewaveform
waveformobtained
obtainedexperimentally
experimentally using
using the
the
anglesin
angles inTable
Table33with
withaamodulation
modulationindex
indexofof0.97.
0.97.

Figure13.
Figure [Link]
Experimentalresult
result
inin a 7-level
a 7-level CHBMLI
CHBMLI using
using the the switching
switching angles
angles found
found for of
for a Mi a Mi of
0.97.
0.97.
Figure 14 shows the table of harmonic values from the fundamental to the 50th
Figure
harmonic. 14observed
It is shows the
thattable of harmonic
the THD values from the fundamental to the 50th
value is 13.61%.
harmonic. It is observed that the THD value is 13.61%.
 Figure 13. Experimental result in a 7-level CHBMLI using the switching angles found for a Mi of
0.97.
Appl. Sci. 2022, 12, 9910 12 of 15
Figure 14 shows the table of harmonic values from the fundamental to the 50th
harmonic. It is observed that the THD value is 13.61%.

Figure14.
Figure [Link]
Harmoniclist
listfor
forthe
theexperimental
experimentalresult
resultin
inaa7-level
7-levelCHBMLI
CHBMLIwith
withMi
Mi==0.97.
0.97.
, 9910 13 of 15
Figure 15
Figure 15 shows
shows the
the output
output voltage
voltage waveform
waveform experimentally
experimentally using
using the
the angles
angles in
in
Table33with
Table withaamodulation
modulationindex
indexofof1.05.
1.05.
Figure 16 shows the table of harmonic values from the fundamental to the 50th
harmonic. It is observed that the THD value is 10.54%.

Figure 15. Experimental result

Figure 15. in a 7-level
Experimental CHBMLI
result in using the
a 7-level CHBMLI switching
using angles
the switching anglesfound fora aMiMi
found for of
of 1.05.
1.05.
Figure 16 shows the table of harmonic values from the fundamental to the 50th
harmonic. It is observed that the THD value is 10.54%.
Appl. Sci. 2022, 12, 9910 13 of 15
Figure 15. Experimental result in a 7-level CHBMLI using the switching angles found for a Mi of
1.05.

Figure16.
Figure [Link]
Harmoniclist
listfor
forthe
theexperimental
experimentalresult
resultin
inaa7-level
7-levelCHBMLI
CHBMLIwith
withMi
Mi==1.05.
1.05.

[Link]
Discussion
In
In recent
recentyears,
years, metaheuristic
metaheuristic methods
methods have
have been
been implemented
implemented to to reduce
reduce THD THD inin
multilevel inverters. Among these methods, the DE algorithm has stood out for
multilevel inverters. Among these methods, the DE algorithm has stood out for its easy its easy im-
plementation
implementation andand
fast fast
convergence; the DE
convergence; thealgorithm has excelled
DE algorithm in several
has excelled research
in several areas.
research
The
areas. following table was created with the data obtained from the simulation (performed
in Matlab
The Simulink)
following and thewas
table experimental tests. the data obtained from the simulation
created with
Table 4 shows the percentage of THD obtained. Fortests.
(performed in Matlab Simulink) and the experimental the Mi of 0.97, the THD obtained
in theTable
simulation
4 shows the percentage of THD obtained. For thethe
was 12.98%, while the THD obtained from Mitests wasthe
of 0.97, 13.61%. A 0.63%
THD obtained
difference between the simulation and experimental percentages.
in the simulation was 12.98%, while the THD obtained from the tests was 13.61%. A 0.63%
difference
Table 4. THDbetween
value (%)the simulation
comparison and experimental
between percentages.
simulation and experimental tests.

Mi Simulation Experimental
0.97 12.98 13.61
1.05 10.46 10.54

For the Mi of 1.05, the THD obtained in the simulation was 10.46%, and the THD in
the tests was 10.54%—a difference of 0.08%. The difference in both THD is negligible.
In both simulation and testing, there is a difference of approximately 3% of THD from
the obtained Mi.

6. Conclusions
The differential evolution algorithm; is recognized as an efficient and robust optimizer.
Usually, differential evolution outperforms its counterparts in efficiency and robustness.
In this paper, the differential evolution algorithm has been analyzed and implemented
in a seven-level cascaded single-phase multilevel inverter (CHBMLI) to obtain the sets of
possible switching angles by varying the modular index (Mi).
The simulation of the differential evolution algorithm implemented in the seven-level
CHBMLI was performed in MATLAB/Simulink software.
The set of angles obtained (without considering the Mi) reduced the THD percentage
to 10.46 (Mi of 1.05).
Appl. Sci. 2022, 12, 9910 14 of 15

Remembering that the Mi must be less than one, it was found that the Mi that satisfies
the established limits and has a decent THD reduction was 0.97, with 12.98% THD.
With these two sets of angles, tests were performed on a seven-level CHBMLI. THDs
of 10.54% (Mi = 1.05) and 13.61% (Mi = 0.97) were obtained.
In both simulation and testing, there was a difference of approximately 3% in THD
from the obtained Mi. The differential evolution algorithm aims to find the optimal angle
sets to reduce THD and provide a suitable modular index for the output load.
In conclusion, effectively minimizing the THD value in an MLI can affect the inverter
output voltage, making the Mi inadequate.
Consideration should be given to whether a high modulus value is required, but not
to a low THD or a low THD with poor modulus value for that load.

Author Contributions: Conceptualization, S.E.D.L.A. and J.A.A.; data curation, J.A.A.; formal anal-
ysis, J.A.A.; funding acquisition, L.G.V.V. and J.D.M.A.; investigation, O.S.V.; methodology, O.S.V.;
project administration, S.E.D.L.A.; resources, L.G.V.V. and J.D.M.A.; software, L.G.V.V. and J.D.M.A.;
supervision, S.E.D.L.A.; validation, O.S.V.; visualization, L.G.V.V. and J.D.M.A.; writing—original
draft, O.S.V. and S.E.D.L.A.; writing—review and editing, J.A.A. All authors have read and agreed to
the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kumar, D.; Gandhi, B.G.R.; Bhattacharjya, R.K. Firefly Algorithm and Its Applications in Engineering Optimization. In Nature-
Inspired Methods for Metaheuristics Optimization, Algorithms and Applications in Science and Engineering; Bennis, F., Bhattacharjya,
R.K., Eds.; Springer: Berlin/Heidelberg, Germany, 2020; Volume 16, p. 503.
2. Qing, A.; Lee, C.K. Differential Evolution in Electromagnetics; Springer: Berlin/Heidelberg, Germany, 2010; Volume 4.
3. Qing, A. Differential Evolution Fundamentals and Applications in Electrical Engineering; John Wiley & Sons (Asia) Pte Ltd.: Hoboken,
NJ, USA, 2009; ISBN 9780470823941.
4. Kabalci, E. Multilevel Inverters Introduction and Emergent Topologies. In Multilevel Inverters; Kabalci, E., Ed.; Academic Press:
Cambridge, MA, USA, 2021; pp. 1–27, ISBN 9780128216682.
5. Upadhyay, D.; Khan, S.A.; Ali, M.; Tariq, M.; Sarwar, A.; Chakrabortty, R.K.; Ryan, M.J. Experimental Validation of Metaheuristic
and Conventional Modulation, and Hysteresis Control of the Dual Boost Nine-Level Inverter. Electronics 2021, 10, 207. [CrossRef]
6. Salam, Z.; Amjad, A.M.; Majed, A. Using Differential Evolution to Solve the Harmonic Elimination Pulse Width Modulation
for Five Level Cascaded Multilevel Voltage Source Inverter. In Proceedings of the 1st International Conference on Artificial
Intelligence, Modelling and Simulation, Kota Kinabalu, Malaysia, 3–5 December 2013; pp. 43–48.
7. Pawar, S.V.; Morteza, S. Harmonic Elimination in Cascade Multilevel Inverter with Non Equal Dc Sources Using Genetic and
Differential Evolution Algorithm. Int. J. Innov. Sci. Eng. Technol. 2015, 2, 144–150.
8. Amjad, A.M.; Salam, Z.; Saif, A.M.A. Application of differential evolution for cascaded multilevel VSI with harmonics elimination
PWM switching. Int. J. Electr. Power Energy Syst. 2015, 64, 447–456. [CrossRef]
9. Chabni, F.; Taleb, R.; Helaimi, M.H. Differential Evolution based SHEPWM for SevenLevel Inverter with Non-Equal DC Sources.
Int. J. Adv. Comput. Sci. Appl. 2016, 7, 304–311. [CrossRef]
10. Jamuna, P.; Rajan, C.C.A. A Heuristic Method: Differential Evolution for Harmonic Reduction in Multilevel Inverter System. Int.
J. Comput. Electr. Eng. 2013, 5, 482–486. [CrossRef]
11. Sudha Letha, S.; Thakur, T.; Kumar, J. Harmonic Elimination in a Solar Powered Cascaded Multilevel Inverter Using Genetic
Algorithm and Differential Evolution Optimization Techniques. In ASME International Mechanical Engineering Congress and
Exposition; American Society of Mechanical Engineers: New York, NY, USA, 2015.
12. Majed, A.; Salam, Z.; Amjad, A.M. Harmonics elimination PWM based direct control for 23-level multilevel distribution
STATCOM using differential evolution algorithm. Electr. Power Syst. Res. 2017, 152, 48–60. [CrossRef]
13. Storn, R.M.; Price, K.V. Differential Evolution—A Simple and Efficient Adaptive Scheme for Global Optimization Over Continuous Spaces;
Technical Report TR-95–012; International Computer Science Institute: Berkeley, CA, USA, 1995.
14. Gutiérrez, D.; López, J.M.; Villa, W.M. Metaheuristic Techniques Applied to the Optimal Reactive Power Dispatch: A Review.
IEEE Lat. Am. Trans. 2016, 14, 11. [CrossRef]
15. Zhang, J.; Sanderson, A.C. Adaptive Differential Evolution: A Robust Approach to Multimodal Problem Optimization (Adaptation,
Learning, and Optimization, 1); Springer: Berlin/Heidelberg, Germany, 2009; Volume 1.
16. Medina, I.R. Algoritmos Bioinspirados: Una Revisión Según sus Fundamentos Biológicos; University of Manchester: Hong Kong,
China, 2014.
Appl. Sci. 2022, 12, 9910 15 of 15

17. Bałchanowski, M.; Boryczka, U. Aggregation of Rankings Using Metaheuristics in Recommendation Systems. Electronics 2022,
11, 369. [CrossRef]
18. Bilal; Pant, M.; Zaheer, H.; Garcia-Hernandez, L.; Abraham, A. Differential Evolution: A review of more than two decades of
research. Eng. Appl. Artif. Intell. 2020, 90, 103479. [CrossRef]
19. Price, K.; Storn, R.; Lampinen, J. Differential Evolution A Practical Approach to Global Optimization; Springer: Alemania, Germany,
2005; p. 542, ISBN 978-3-540-31306-9.
20. Malik, H.; Iqbal, A.; Joshi, P.; Agrawal, S.; Bakhsh, F.I. Metaheuristic and Evolutionary Computation: Algorithms and Applications;
Springer Nature: Singapore, 2021; p. 830, ISBN 978-981-15-7571-6.
21. Montes, E.M. Paradigmas emergentes en algoritmos bio-inspirados. In Inteligencia Aritificial; Alfaomega, Ed.; Alfaomega: Mexico
City, Mexico, 2006; pp. 504–533.
22. Ronkkonen, J.; Kukkonen, S.; Price, K.V. Real-Parameter Optimization with Differential Evolution. In Proceedings of the IEEE
Congress on Evolutionary Computation, Edinburgh, UK, 2–5 September 2005; p. 8.
23. Castillo, E.J. Esquema Adaptativo para el Manejo de Restricciones de Límite en Problemas de Optimización Numérica Restringida.
Ph.D. Thesis, Centro de Investigación en Inteligencia Artificial Universidad Veracruzana Xalapa, Veracruz, Mexico, 2019.
24. Juárez-Castillo, E.; Pérez-Castro, N.; Mezura-Montes, E. An Improved Centroid-Based Boundary Constraint-Handling Method in
Differential Evolution for Constrained Optimization. Int. J. Pattern Recognit. Artif. Intell. 2017, 31, 1759023. [CrossRef]
25. Yong, J. Optimization Theory a Concise Introduction; World Scientific Publishing Co. Pte. Ltd.: Singapore, 2018; ISBN 9813237643.
26. Chong, E.K.P.; Zak, S.H. An Introduction To Optimization, 4th ed.; John Wiley & Sons, Inc. Publication: Hoboken, NJ, USA, 2013;
ISBN 978-1-118-27901-4.
27. Sánchez Vargas, O.; De León Aldaco, S.E.; Aguayo Alquicira, J.; López Núñez, A.R. Evolutionary Metaheuristic Methods Applied
to Minimize the THD in Inverters: A Systematic Review. Eur. J. Electr. Eng. 2021, 23, 237–245. [CrossRef]
28. Hamzah, H.H.; Ponniran, A.; Kasiran, A.N.; Harimon, M.A.; Gendum, D.A.; Yatim, M.H. A Single Phase 7-Level Cascade Inverter
Topology with Reduced Number of Switches on Resistive Load by Using PWM. J. Phys. Conf. Ser. 2018, 995, 012061. [CrossRef]
29. Siddiqui, N.I.; Alam, A.; Quayyoom, L.; Sarwar, A.; Tariq, M.; Vahedi, H.; Ahmad, S.; Mohamed, A.S.N. Artificial Jelly-
fish Search Algorithm-Based Selective Harmonic Elimination in a Cascaded H-Bridge Multilevel Inverter. Electronics 2021,
10, 1402. [CrossRef]
30. Wei, S.; Wu, B.; Li, F.; Sun, X. Control Method for Cascaded H-Bridge Multilevel Inverter with Faulty Power Cells. In Proceedings
of the Eighteenth Annual IEEE Applied Power Electronics Conference and Exposition., Miami Beach, FL, USA, 9–13 February 2003;
pp. 261–267.
31. Sánchez Vargas, O.S.; De León Aldaco, S.E.; Aguayo Alquicira, J.; Flores Rodríguez, E.; Lozoya Ponce, R.E. Cálculo de los ángulos
óptimos de conmutación para un inversor multinivel utilizando evolución diferencial. Pist. Educ. 2022, 43, 141.