energies

Article
Influence of a Hybrid MPPT Technique, SA-P&O, on PV
System Performance under Partial Shading Conditions
Ahmed G. Abo-Khalil 1 , Ibrahim I. El-Sharkawy 1,2 , Ali Radwan 1,2, * and Saim Memon 3

1 Department of Sustainable and Renewable Energy Engineering, University of Sharjah,

Sharjah P.O. Box 27272, United Arab Emirates
2 Mechanical Power Engineering Department, Mansoura University, Mansoura 35516, Egypt
3 School of Engineering, Arden University, Arden House, Middlemarch Park, Coventry CV3 4FJ, UK
* Correspondence: [email protected]; Tel.: +971-501595571

Abstract: The electricity sector has been undergoing profound transformations. In particular, the
Portuguese self-consumer regime has allowed customers of the medium and low voltage electricity
grid to be producers/consumers of electricity, actively contributing to greater energy efficiency. In
this context, the energy that comes from the sun is not used to its maximum. In addition, photovoltaic
cells have a characteristic operating curve (voltage vs. current), in which any operating point is
reflected. Within this curve, there is a particular point known as the maximum power point (MPP) at
which the cell supplies the maximum power output to a load. If the cell does not operate at this point,
it has lower efficiency values. To harness maximum power under standard and dynamic shading
conditions, there are various techniques of low complexity for capturing maximum power. We present
a maximum power point tracking (MPPT) algorithm capable of dealing with the problem of partial
shading. This algorithm involves modifying one of the most used algorithms within photovoltaic
systems, known as P&O, using a simulated annealing (SA) algorithm. P&O is often used due to
its straightforward implementation, but it is susceptible to partial shade conditions. Sampling was
added to this algorithm to a better approach to the point of maximum power using the SA, and then
to attain a more precise convergence with P&O. Implementing a maximum power point tracking
method under partial shading was the major goal of this study.
Citation: Abo-Khalil, A.G.;
El-Sharkawy, I.I.; Radwan, A.; Keywords: global maximum power point (GMPP); local maximum power point (LMPP); simulated
Memon, S. Influence of a Hybrid annealing (SA); perturbation and observation (P&O)
MPPT Technique, SA-P&O, on PV
System Performance under Partial
Shading Conditions. Energies 2023,
16, 577. https://doi.org/10.3390/
1. Introduction
en16020577
Traditional energy sources, such as fossil fuels and nuclear energy, have environmental,
Academic Editor: Manolis Souliotis social and political problems because they are highly polluting. Due to the constant energy
Received: 25 November 2022
demand of the industrial sector and the increase in the use of electrical equipment in
Revised: 16 December 2022 a variety of human activities, the demand for electrical energy has increased exponentially
Accepted: 25 December 2022 in recent decades. Renewable energy sources are sustainable alternatives to meet the need
Published: 4 January 2023 for electricity, provide the diversification of the matrix and increase energy security for
countries in terms of supply. Options for generating renewable energy include hydraulic,
biomass, wind, and solar [1]. Among renewable sources, solar energy has great potential to
contribute to the world’s demand for electricity, especially in countries with high levels of
Copyright: © 2023 by the authors. solar radiation. The world’s solar energy power in 2015 was 227 GW, and the prospect is to
Licensee MDPI, Basel, Switzerland. reach a power of 1362 GW by 2030,. This source encompasses large solar plants or parks,
This article is an open access article and distributed generation on the roofs of buildings [2].
distributed under the terms and Photovoltaic cells have characteristic V-I curves that define their behavior under
conditions of the Creative Commons
different operating conditions. The power of a solar cell is given by the product of the
Attribution (CC BY) license (https://
current and the voltage of the cell. The maximum power point (MPP) is the product of the
creativecommons.org/licenses/by/
voltage at the maximum point (VMPP ) and current at the maximum point (IMPP ) for which
4.0/).

Energies 2023, 16, 577. https://doi.org/10.3390/en16020577 https://www.mdpi.com/journal/energies

Energies 2023, 16, 577 2 of 17

the power extracted from the photovoltaic system is maximum (PMPP ). The maximum
power point varies continuously, depending on factors such as the temperature of the
solar cell and the irradiance conditions. To increase the competitiveness of PV energy with
respect to other types, it is necessary to increase its efficiency, which is low, or reduce the
costs. The MPP changes over time since it depends on the temperature and the level of
irradiance; therefore, to make the photovoltaic modules operate close to the MPP point, the
use of tracking algorithms is needed to be linked to a power converter, which is usually
a direct current to direct current (DC/DC) device. The photovoltaic module then delivers
specific current and voltage values that maximize power delivery.
To deal with nonlinear phenomena and achieve system stability with MPPT, many
approaches have been proposed in the literature, such as techniques with adapted con-
stant voltage [3], fuzzy logic [4–6], neural Networks [7,8], perturbation and observation
(P&O) [9–12], and incremental conductance [13–15].
The choice of technique depends on factors such as speed and accuracy in tracking,
the level of complexity of the sensors used, and the cost of the necessary equipment. Con-
ventional techniques, such as perturbation and observation, and incremental conductance,
have easy implementation. Techniques based on artificial intelligence, such as particle
swarm optimization, and ant colony optimization, as well as techniques based on fuzzy
logic and artificial neural networks, have a greater degree of complexity and require greater
computational effort [16].
Regarding efficiency, two situations must be analyzed. When the system is subjected
to homogeneous levels of irradiance, when all its modules receive the same value of solar
power, conventional techniques are able to track the MPPT efficiently. However, when the
modules receive different levels of irradiance in a condition known as partial shading or
PSC (partial shading condition), the PV curve, due to the construction of the modules, has
more than one power peak, one of which is a global peak and the others, with lower power
values, are known as local peaks [17].
In this condition, conventional techniques track the first peak found, which can be
a global or a local one, resulting in a loss of power if tracking a local peak. Therefore, under
PSC, it is preferable to use techniques capable of distinguishing the power level between
the different peaks. In this case, the best examples are based on artificial intelligent concepts.
Several algorithms are used to track the maximum power point, as already mentioned
when they come across the LMPP points, and PSO (Particle Swarm Optimization), ACO
(Ant Colony Optimizer), and Gray Wolf Optimization (GWO) algorithm are used to solve
the problems associated with partial shading [18–20].
In our research, new hybrid techniques were implemented experimentally and theo-
retically. The proposed optimization system, based on a meta-heuristic SA, conducts the
operation of the boost converter to GMPP promptly and accurately under any of the shad-
ing conditions. To perform the MPPT, the SA algorithm is used separately. Its combination
with the P&O, initiating the SA-P&O algorithms, boosts converter operation to GMPP
under any of the conditions. To evaluate the performance of the SA and SA-P&O-based
MPPT algorithms, a comparison was made with the traditional MPPT algorithm, i.e., the
P&O method. The implemented MPPT techniques were compared and evaluated consider-
ing convergence time, the amplitude of the steady-state oscillations in the MPP, and the
methods’ efficiencies during uniform irradiance and partial shading.

2. System Description
Figure 1 shows the PV system used in this study. The 60 Wp PV panel is connected by
the DC-DC boost converter and a single-phase inverter for connection to the grid. A Voltage
Source Inverter (VSI) is used to perform the DC/AC conversion, and a connection filter of
the L type is used to connect the VSI to the grid.
 erence current provided by the algorithm for operating the arrangement at maximum
power, which is compared with the operating current of the PV panel. The result of this
comparison
2. constitutes the input of a proportional integral (PI) controller that produces a
System Description
modulation
Figure 1index
showsofthe
thePV
PWM connected
system used intothis
thestudy.
switching device
The 60 Wp PVof the boost
panel converter
is connected
connected to the photovoltaic array. In the double-stage PV system used
by the DC-DC boost converter and a single-phase inverter for connection to the in this work,
grid.the
Energies 2023, 16, 577 3 of 17A
single-phase DC-AC inverter is responsible for both the voltage control of the
Voltage Source Inverter (VSI) is used to perform the DC/AC conversion, and a connectionDC bus and
the current
filter injected
of the L type isinto
usedthe
to grid.
connect the VSI to the grid.
The block diagram of the MPPT method is shown in Figure 1, where Ipv and Vpv are
PV Array DC-DC Converter DC-AC Inverter
the current and voltage measurements
Lg of the photovoltaic panel, respectively. Vref is a ref-
erence current provided by the algorithm for operating the arrangement at maximum
power,+which is compared with thee operating current of the PV panel. The result of this
S bc Cdc V g
dc constitutes the input of a proportional integral (PI) controller that produces a
comparison
- S inv converter
modulation index of the PWM ig connected to the switching device of the boost
V *
connected to the photovoltaic array. In the double-stage i *
PV system v P
used in this work, the
dc g
eg +
single-phase DC-AC inverter is responsible PI inv of
PI voltage control
for both the W the DC bus and
Ipv - M
the current injected into the grid. i V sinθg v
Vd dcg g
PLL
Vpv c
PV Array DC-DC Converter DC-AC Inverter V S
dc V ref bc
Lg Ipv MPPT + PI PWM
Vpv V
pv -
+ eg
S bc Cdc V
dc
- S inv
ig
Figure 1. PV grid-connected system. *
Vdc ig* v P
PI inv W
Figure 1. PV grid-connected system. eg + PI
Ipv 2.1. PV Model M
The block diagram of the MPPT method ig Vis- shown in Figure 1, vwhere Ipv and Vpv
sinθg
d dc
are theSolar cellsand V made of a semiconductor material, usually PLL
are silicon g(Si). Each cell has a
Vpv current cvoltage measurements of the photovoltaic panel, respectively. Vref is
athin layer of
reference type nprovided
current material by
andthe
anotherV withfor
algorithm a greater
V ref thickness
operating Sat maximum
of type p material,
the arrangement form-
dc bc
ing thewhich
power, p-n junction. Whenwith
is compared exposed toIpv
the operating MPPT
photons from
current
+ solar
of the PI
radiation,
PV PWM
panel. the electrons
The result are
of en-
this
ergized, creating free charge carriers V
that move between V the bands of the material, causing
pv
comparison constitutes the input of a proportional pv -
integral (PI) controller that produces
aamodulation
potential difference
index of between
the PWMthe cell terminals
connected to thefrom the electric
switching devicefield, generating
of the an elec-
boost converter
tric current.
connected to The photocurrentarray.
the photovoltaic effectInis represented by a current
the double-stage source,
PV system usedasinshown in Figure
this work, the
2, where
single-phase
Figure Rshgrid-connected
1. PV and RS represent
DC-AC inverter isthe parallel and
responsible
system. the series
for both resistance,
the voltage respectively.
control Bothand
of the DC bus are
intrinsic
the current elements
injectedof thethe
into cellgrid.
[21].
2.1. PV Model
2.1. PV Model
R
Solar cells are made of a semiconductor material, usually silicon (Si). Each cell has a
s I
thinSolar
layer cells arenmade
of type of aand
material semiconductor
another withmaterial,
a greater usually silicon (Si). Each cell has
L thickness of type p material, form-
aing
thin layer of type n material and another with a greater
the p-n junction. When exposed to photons from solar radiation, thickness the
of type p material,
electrons are en-
I +
forming the p-n junction. When exposed to photons from solar radiation, the electrons
areph I
ergized, creating free charge carriers that move between the bands of the material, causing
energized, creating free d charge
a potential difference between
carriers that move between the bands of the material,
the cell terminals from the electric field, generating an elec-
causing a potential difference between the cell terminals from the electric field, generating
tric current. The photocurrent effect is represented by a current source, as shown in Figure
an electric current. The photocurrent effect is represented by a current source, as shown
D R V
2, where Rsh and RS represent the parallel and the series resistance, respectively. Both are
in Figure 2, where Rsh and RS represent sh
intrinsic elements of the cell [21].
the parallel and Lthe series resistance, respectively.
Both are intrinsic elements of the cell [21].

Rs I
L-

Iph +
Id

D Rsh VL

-
Figure 2. PV equivalent circuit [21].
 𝐴.𝑉𝑡 𝑅𝑠ℎ

where I0 is the leakage current or reverse saturation of the diode [A], q is the electron
charge (1.60217646 × 10−19 C), k is the Boltzmann constant (1.3806503 × 10−23 J/K), T is the
junction temperature p-n [K], A is the diode ideality factor (1 ≤ to ≤ 2), NS is the number
Energies 2023, 16, 577
of cells connected in series, and Vt is the panel thermal voltage given by: 4 of 17
𝐾.𝑇
𝑉𝑡 = (2)
𝑞

The When the of

behavior solar cell is irradiated
a photovoltaic panel with light,described
is thus an equivalent
by twocircuit of the solarcurves,
characteristic cell can be
drawn, as shown
current-voltage in Figure 2. At this
and power-voltage. These time,
curvesthe output current
are usually of the solar
available fromcell theisPVas man-
shown in
Equation
ufacturer and (1)
are[22]:
of great importance because their values change according to the tem-
perature and radiation conditions.  The  characteristics  of the voltage-power and  voltage-
current characteristic q ( V L + I L · R s ) V L + Ns · I L · R s
IL =curves
I ph − of
Io aexp
photovoltaic module−are 1 shown,
− respectively, in Figure (1)
3a,b for radiation of 1000 W/m2 and a temperature A·Vt of 25 °C. Rsh
It is essential
where Io is thetoleakage
highlight the five
current ormain
reversepoints of the characteristic
saturation of the diode curves:
[A], q isthe theopen-
electron
circuit voltage V oc; the short-circuit
charge (1.60217646 × 10 − 19 current I sc; maximum power Pmax, and the maximum
C), k is the Boltzmann constant (1.3806503 × 10−23 J/K), T
voltage
is theand currenttemperature
junction values related p-n to Pmax
[K], A ,isVthemax, and Imax. The appearance of the graph
diode ideality factor (1 ≤ to ≤ 2), NS is the
appearance
number of cells connected in series, and Vt is theirradiation.
in Figure 3 varies with temperature and On cloudy
panel thermal voltagedays,giventheby:irra-
diation is low, and the current produced is proportionately low. For days when the tem-
perature is higher, the voltage decreases proportionately. K·T Equations (3) and (4) show the
Vt = (2)
dependency of the open circuit voltage on the PV open q circuit voltage.
𝐾.𝑇 𝐽𝑠𝑐
The behavior of a photovoltaic 𝑉𝑜𝑐 = panel lnis(thus+described
1) by two characteristic curves,
(3)
current-voltage and power-voltage. These curves 𝑞 𝐽0 are usually available from the PV manu-
facturer and are of great𝑑𝑉importance
𝑜𝑐 𝑉 𝐾.𝑇 their
because 1 𝑑𝐽𝑠𝑐
values1 𝑑𝐽
change according to the tempera-
= ( 𝑜𝑐) + ( − 0
) (4)
ture and radiation conditions.𝑑𝑇 The𝑇 characteristics
𝑞 𝐽𝑠𝑐 𝑑𝑡 of the𝐽voltage-power
0 𝑇 and voltage-current
characteristic curves of a photovoltaic module are shown, respectively, in Figure 3a,b for
where, Jsc = Jph, Voc is related2 to Jsc and Jo, and hence to ◦the bandgap (Eg).
radiation of 1000 W/m and a temperature of 25 C.

Impp

(a)

Pmpp

Vmpp

(b)

Figure 3. Characteristic curves of a photovoltaic module: (a) voltage-current curve, (b) voltage-
power curve.

It is essential to highlight the five main points of the characteristic curves: the open-
circuit voltage Voc ; the short-circuit current Isc ; maximum power Pmax , and the maximum
voltage and current values related to Pmax , V max , and Imax . The appearance of the graph
appearance in Figure 3 varies with temperature and irradiation. On cloudy days, the
irradiation is low, and the current produced is proportionately low. For days when the
temperature is higher, the voltage decreases proportionately. Equations (3) and (4) show
the dependency of the open circuit voltage on the PV open circuit voltage.

K·T
 
Jsc
Voc = ln +1 (3)
q Jo
Energies 2023, 16, 577 5 of 17

K·T
   
dVoc Voc 1 dJsc 1 dJo
= + − (4)
dT T q Jsc dt Jo T
where, Jsc = Jph , Voc is related to Jsc and Jo , and hence to the bandgap (Eg ).
MPPT methods can be classified depending on different factors or characteristics
that differentiate them from one another, such as their implementation complexity or
convergence speed, which cause their efficiency to vary. There are a large number of
characteristics by which methods can be classified. Table 1 shows the main maximum
power point monitoring methods according to their different characteristics.

Table 1. Main characteristics of the maximum power point tracking methods.

Analog Setting Speed of Implementation Parameters Method

Method Cost
or Digital Parameters Convergence Complexity Sensed Type
Voltage
P&O Both of them No It varies High Direct High
and current
Voltage
Inc Conductance Digital No It varies High Direct High
and current
Ripple
Voltage
Correlation Analog Yes Fast High Direct High
and current
Control
Maximization of
voltage or Analog No Fast Low Voltage Direct Low
current on load
Voltage
Curve fitting Digital Yes Varies Low Indirect Low
and current
Voltage
Look-up table Digital Yes Average Average Indirect Low
and current
Factional open Voltage
Both Yes Average Average Indirect Low
circuit voltage and current
Factional short Voltage
Both Yes Average Average Indirect Low
circuit voltage and current
Voltage
Fuzzy logic Digital Yes Fast High Both High
and current
Voltage
Neural Network Digital Yes Fast Average Both High
and current
PSO, ACO, Voltage
Digital Yes Fast High Both High
GWO, SA. and current

2.2. Shading Effect

Throughout the day, variations in irradiation and temperature occur naturally. There
are also situations, such as total or partial shading on the module due to clouds, trees,
leaves, and buildings that cause sudden variations in these levels, as shown in Figure 4.
This causes power reduction. As the cells are usually connected in series, all modules are
affected, in addition to causing the risk of damage to the modules due to the so-called
hotspot that can lead to rupture of the glass and melting of polymers that are part of the
module structure [23].
Continuing with the model of the equivalent circuit of the photovoltaic panel, the effect
that partial shading has in photovoltaic systems is considered below. For this, four panels
connected in series are shown (Figure 5). The four panels, seen as an equivalent circuit, are
also shown in Figure 6. A photovoltaic module is partially shaded when it does not receive
the total available irradiation in one or more modules in the chain. Some causes of this
shading are trees, birds, snow, and buildings, among others.
 Energies
Energies 2023,
2023, 16,
16,16, 577 PEER REVIEW
x FOR 6 of
6 6of 17
1818
Energies
Energies 2023,
2023, 16, xx FOR
FOR PEER
PEER REVIEW
REVIEW 6 of
of 18

Figure 4.4.
Figure
Figure PV
4. under
PV
PV partial
under
under shading.
partial
partial shading.
shading.
Figure 4. PV under partial shading.

(-) (-)
(-) (+)(+)
(+)
(-) (-)
(-) (+)(+) (-) (-) (+)(+) (-) (-) (+)(+) (-) (-) (+)(+)
(+) (-) (+) (-) (+) (-) (+)

Figure 5.5.
Figure
Figure Series
5. connection
Series
Series ofof
connection
connection photovoltaic
of panels.
photovoltaic
photovoltaic panels.
panels.
Figure 5. Series connection of photovoltaic panels.
Iph IIph Iph IIph Iph IIph Iph IIph
ph ph ph ph

RshR
1R
sh1
RshR
2R
sh2
RshR
3R
sh3
RshR
4R
sh4
sh1 sh2 sh3 sh4
Rs1RRs1 Rs2RRs2 Rs3RRs3 Rs4RRs4
s1 s2 s3 s4

(-) (-)
(-) (+) (+)
(+)

Figure 6.6.
Figure
Figure
Figure Equivalent
6.
6. circuit
Equivalent
Equivalent
Equivalent ofof
circuit
circuit
circuit photovoltaic
of
of panels.
photovoltaic
photovoltaic
photovoltaic panels.
panels.
panels.

IfIf
Ifwewetake
we takethe
take theexample
the exampleofof
example ofFigure
Figure6 66and
Figure andshade
and shadea aapanel
shade panelofof
panel ofthe thearray
the array(Figure
array (Figure7),7),
(Figure 7),the
the
the
shadow
shadow
shadow atatmodule
at module
module 2 2does not
2 does
does notallow
not allow
allow thethe
thesupply
supply
supply ofof
current,
of current,
current,and
and
andthere
there
there isis
isa alimitation
a limitation
limitation inin
initsits
its
supply,
supply,
supply,
supply, soso
the
so theequivalent
the equivalent
equivalent circuit,
circuit,inin
circuit, this
in thiscase,
this would
case,
case, would
would
would bebe
be
bethe
theone
the
the one
oneshown
one shown
shown
shown inin
inFigure
in Figure
Figure
Figure8.8.
8.
8.
 Energies 2023, 16, x FOR PEER REVIEW 7 of 18
Energies2023,
Energies
Energies 2023,16,
2023, 16,x577
16, xFOR
FORPEER
PEERREVIEW
REVIEW 777ofof 18
of18
17

Figure 7. Partial shading in photovoltaic panels.

Figure7.7.
Figure
Figure 7.Partial
Partialshading
Partial shadingin
shading inphotovoltaic
in photovoltaicpanels.
photovoltaic panels.
panels.

Figure 8. Equivalent circuit of photovoltaic panels with partial shading.

Figure8.8.
Figure
Figure 8.Equivalent
Equivalentcircuit
Equivalent circuitof
circuit ofphotovoltaic
of photovoltaicpanels
photovoltaic panelswith
panels withpartial
with partialshading.
partial shading.
shading.
It
ItIt can
can
Itcan
canbebe
be seen
seen
beseen that
that
seenthat because
because
thatbecause the
the
becausethe source
source
thesource does
does
sourcedoes
doesnotnot
not receive
receive
notreceive radiation,
radiation, itdoes
radiation,itit
receiveradiation, does
does not
not
doesnot supply
supply
notsupply
supply
current
current
currentand
current and behaves
andbehaves
behaves like
behaveslike
like an
likeanan
an open
open
open
open circuit,
circuit,
circuit, so
circuit,
soso
so the
the current
thethe
current
current flows
current flows
flows
flows through
through
through
through the resistance
the RR
resistance
theresistance
the resistance sh2
Rsh2
sh2
(Figure
R
(Figure
(Figure 9).
(Figure
9). 9).
9).
sh2

Figure 9.
9. Current circulation
circulation when
when partial
partial shading
shading occurs.
occurs.
Figure9.9.Current
Figure Currentcirculation
circulationwhen
whenpartial
partialshading
shadingoccurs.
occurs.

In Figure 10, aa characteristic

characteristic curve is represented when some of a module’s module’s cells are
InFigure
In Figure10,
10,aacharacteristic
characteristiccurve
curveisisrepresented
representedwhen whensome someof ofaamodule’s
module’scells
cellsare
are
shaded. The algorithm needs to be robust and efficient to be able to to detect
detect the
the maximum
maximum
shaded.The
shaded. Thealgorithm
algorithmneeds needsto tobe
berobust
robustand andefficient
efficientto tobebeable
abletotodetect
detectthe
themaximum
maximum
power
power point and
andprevent
prevent itfrom frombeing
being intop
a top call place. Due to partial shading, the
powerpoint
power pointand preventitit
andprevent itfrom
frombeing
beinginin
inaaatop call
topcall place.
callplace.
place.DueDue
Dueto to partial
topartial shading,
partialshading, the
shading,the mod-
themod-
mod-
modules
ules have have different
different IV IV curves.
curves. WhenWhen
they they
are are connected
connected in in series
series or or parallel,
parallel, different
different IV
uleshave
ules havedifferent
differentIV IVcurves.
curves.WhenWhenthey
theyareareconnected
connectedin inseries
seriesororparallel,
parallel,different
differentIVIV
IV curves
curves areare superimposed,
superimposed, generating
generating ananIVIVcurve
curvewith
withmultiple
multipleinflections
inflections and,
and, conse-
conse-
curvesare
curves aresuperimposed,
superimposed,generating
generatingan anIV IVcurve
curvewith
withmultiple
multipleinflections
inflectionsand,
and,conse-
conse-
quently,
quently, a PV curve with various maximum power points, as shown in Figure 10. Under
quently,aaaPV
quently, PV curve
PVcurve
curvewith with various
withvarious maximum
variousmaximum
maximumpower power points,
powerpoints,
points,as as shown
asshown
shownin in Figure
inFigure 10.
Figure10. Under
10.Under
Under
the
the effect
effect of
of PSC,
PSC, the
the system
system has
has a considerable loss of control, which can become even
theeffect
the effectofofPSC,
PSC,the thesystem
systemhas hasaaaconsiderable
considerable
considerableloss loss
lossofof control,
ofcontrol, which
control,which
whichcan can become
canbecome
becomeeveneven
even
more
more significant
significant if
if it
it does
does not
not have
have a
a proper
proper MPPT
MPPT method.
method.
moresignificant
more significantififititdoes
doesnot
nothave
haveaaproper
properMPPTMPPTmethod.
method.
 Energies2023,
Energies 16,x577
2023,16, FOR PEER REVIEW 8 of1817
8 of

Figure 10. PV V-I and V-P characteristics under partial shading.

Figure 10. PV V-I and V-P characteristics under partial shading.
3. MPPT Using Simulated Annealing
3. MPPT Using Simulated Annealing
The simulated annealing method in Figure 11 is based on a metallurgy technique, and
Energies 2023, 16, x FOR PEER REVIEW The simulated
involves controlled annealing method
heating and coolinginofFigure 11 isThis
a material. based on a metallurgy
technique oftechnique,
aims to 9reduce18 defects
and involves controlled heating and cooling
by increasing the size of a material’s crystals.of a material. This technique aims to reduce
defects by increasing the size of a material’s crystals.
The heating and cooling of the material affect the temperature and the free thermo-
dynamic energy. Although the same decrease in temperature causes the same decrease in
free thermodynamic energy, cooling at a given rate causes a more significant reduction in
free energy.
This temperature reduction approach is implemented in the simulated annealing al-
gorithm as it is the optimum solution [24–26].
The method starts with a random solution, Pi and selects a neighboring solution Pk.
After calculating the cost of the two solutions, their difference ∆P = Pk − Pi is calculated. If
∆P is less than zero, then the value of the objective function of the neighboring solution is
better, and replaces the current solution Pi with Pk.
If ∆P is greater than or equal to zero, the solution with the worst result is accepted.
The value of this probability decreases with the number of iterations. If the value of pr is
more significant than a random number between zero and one, the solution with the worst
result is accepted. This procedure is repeated until a stopping criterion is satisfied.
One of the main advantages of this algorithm is that it allows testing solutions that
differ from those currently available. It thus minimizes the possibility of the energy gen-
erated by the system oscillating over a local maximum point indefinitely.
To extract the MPPT of a PV module, the probability of acceptance is given by [27]:

𝑃 = 𝑒𝑥𝑝 (5)

where Pk is the measured power, Pi is the best power defined up to the time of the search,
and the parameter Tk corresponds to the current variable parameter with iterations (called
temperature) of the search process. Several methods can define parameter reduction; how-
ever, there are two commonly used methods. One consists of geometric reduction, which
is given by [28,29]:
𝑇 = 𝛼 𝑇(𝑘 − 1) (6)
where α < 1 is the geometric cooling constant 𝑇(𝑘 − 1). Another method for updating the
parameter Tk was proposed by [30], and calculated as follows:
( )
𝑇 = (7)
( )

where β controls Lundy’s cooling schedule; however, regardless of how Tk is updated, this
method needs to be restarted when changes in the conditions of the environment occur.
The first equation for Tk is best used since there is no need to restart the calculations in the
case of environmental
Figure 11. Flowchart of thechanges.
SA technique.
Figure 11. Flowchart of the SA technique.
3.1. Perturbation and Observation (P&O) Algorithm
The P&O method is based on the modification of the trip ratio or duty cycle (D) of
the boost converter that modifies the value of the current supplied by the photovoltaic
panel. This method consists of monitoring the board operating voltage (Vpv); for example,
Energies 2023, 16, 577 9 of 17

The heating and cooling of the material affect the temperature and the free thermo-
dynamic energy. Although the same decrease in temperature causes the same decrease in
free thermodynamic energy, cooling at a given rate causes a more significant reduction in
free energy.
This temperature reduction approach is implemented in the simulated annealing
algorithm as it is the optimum solution [24–26].
The method starts with a random solution, Pi and selects a neighboring solution Pk .
After calculating the cost of the two solutions, their difference ∆P = Pk − Pi is calculated. If
∆P is less than zero, then the value of the objective function of the neighboring solution is
better, and replaces the current solution Pi with Pk .
If ∆P is greater than or equal to zero, the solution with the worst result is accepted.
The value of this probability decreases with the number of iterations. If the value of pr is
more significant than a random number between zero and one, the solution with the worst
result is accepted. This procedure is repeated until a stopping criterion is satisfied.
One of the main advantages of this algorithm is that it allows testing solutions that
differ from those currently available. It thus minimizes the possibility of the energy
generated by the system oscillating over a local maximum point indefinitely.
To extract the MPPT of a PV module, the probability of acceptance is given by [27]:

Pk − Pi
 
Pr = exp (5)
Tk

where Pk is the measured power, Pi is the best power defined up to the time of the search,
and the parameter Tk corresponds to the current variable parameter with iterations (called
temperature) of the search process. Several methods can define parameter reduction;
however, there are two commonly used methods. One consists of geometric reduction,
which is given by [28,29]:
Tk = α T (k − 1) (6)
where α < 1 is the geometric cooling constant T (k − 1). Another method for updating the
parameter Tk was proposed by [30], and calculated as follows:

T ( k − 1)
Tk = (7)
1 − β T ( k − 1)

where β controls Lundy’s cooling schedule; however, regardless of how Tk is updated, this
method needs to be restarted when changes in the conditions of the environment occur.
The first equation for Tk is best used since there is no need to restart the calculations in the
case of environmental changes.

3.1. Perturbation and Observation (P&O) Algorithm

The P&O method is based on the modification of the trip ratio or duty cycle (D) of
the boost converter that modifies the value of the current supplied by the photovoltaic
panel. This method consists of monitoring the board operating voltage (Vpv); for example,
if the voltage rises (disturbance) and the power drawn by the load increases (observation)
it means that the operating point has moved towards the maximum power point, at which
point the operating voltage should be disturbed in the same direction. By contrast, if the
power extracted by the load decreases, the operating point has moved in the opposite
direction to the maximum power point as evidenced in the flow diagram in Figure 12.
Energies 2023,16,
Energies2023, 16,577
x FOR PEER REVIEW 10 of
10 of 17
18

Energies 2023, 16, x FOR PEER REVIEW 10 of 18

Figure 12.
Figure12.
Figure Flowchart
12.Flowchart ofthe
Flowchartof
of theP&O
the P&Otechnique.
P&O technique.
technique.

3.2.
3.2.SA-P&O
3.2. SA-P
SA-P &&O Algorithm
O Algorithm
Algorithm
The
TheSA-P&O
The SA-P
SA-P & O algorithm
algorithmuses
algorithm usesthe
uses theSA
the SAtechnique
SA techniqueto
technique totoreachaapoint
reach
reach apoint
point closeto to
close
close tothe
the the global
global
global
maximum
maximumpower
maximum powerpoint.
power point. Soon
point. after
Soonafter
Soon finding
afterfinding it,
findingit, the
it,the algorithm
thealgorithm
algorithm switches
to to
switches
switches to the
the the
P&OP&O
P&O method,
method,
method,
which
whichstarts
which starts with
starts referencingthe
with referencing
referencing thebest
the best
best power
power
power point point
point foundfound
found defined
defined theby
byby
defined SAthe
the SA method.
method.
SA method.Fig-Fig-
Figure
ure13
ure 13 shows
13shows
shows the the flowchart
the flowchart
flowchart for for this
forthis algorithm,
thisalgorithm,
algorithm,the the
theSA-PSA-P&O
SA-P && OO MPPT. Figure
MPPT. Figure
Figure 13 shows
13 shows
13 showsthethe
the
SA-P&O
SA-P&Ohybrid
SA-P&O hybridMPPT
hybrid MPPTalgorithm.
algorithm.
algorithm.

Figure 13. Flowchart of the SA-P&O technique.

Figure 13. Flowchart of the SA-P&O technique.
Figure 13. Flowchart of the SA-P&O technique.
Energies 2023, 16, x FOR PEER REVIEW 11 of 18
Energies 2023,
Energies 2023, 16,
16, 577
x FOR PEER REVIEW 1111of
of 18
17

4. Results and Discussion

4.
4. Results
Results andand Discussion
Discussion
The performance of the SA-P&O-based MPPT algorithm was evaluated using exper-
The performance
Theresults
performance of the
the SA-P&O-based MPPT algorithm
algorithm was evaluated
evaluated using
using experi-
exper-
imental for theofsystem SA-P&O-based
configurationMPPT shown in Figurewas 14. An empirical analysis
imental
mental results
results for for the system
the system configuration
configuration shown in Figure
shown in Figure of 14.
14.the An
AnMPPT, empirical
empirical analysis
analysis was
was performed using Microlabbox for the implementation a DC-DC boost
was performed
performed using using Microlabbox
Microlabbox for forimplementation
the the implementation of ofMPPT,
the the MPPT,
a a DC-DC
DC-DC boost boost
con-
converter, a DC/AC converter to adapt the voltage to the connection grid utility, and a
converter,
verter, aLDC/AC a DC/AC converter
converter to adapt
to adapt the voltage
the voltage to theto the connection
connection grid and
gridcompared
utility, utility, and a
passive filter were used. The SA-P&O-based MPPT algorithm was toa the
passive
tra-
passive
L filter L filter
were were
used. Theused. The SA-P&O-based
SA-P&O-based MPPT MPPT algorithm
algorithm was was compared
compared to the to the tra-
traditional
ditional P & O-based MPPT algorithm. Both MPPT algorithms were compared consider-
ditional
P &the P & O-based
O-based MPPT MPPT algorithm.
algorithm. Both MPPT Both algorithms
MPPT algorithmswere were compared
compared consider-
considering
ing following cases: (1) the photovoltaic arrangement operates in STC (as shownthe in
ing the following
following cases: cases:
(1) the (1) the photovoltaic
photovoltaic arrangement
arrangement operates operates
in STC inshown
(as STC (as in shown
Figure in
5),
Figure 5), and case (2) when the photovoltaic arrangement works under partial shading
Figure 5), and case (2) when the photovoltaic arrangement works
and case (2) when the photovoltaic arrangement works under partial shading conditions under partial shading
conditions (as shown in Figure 7). In the latter case, two photovoltaic panels were partially
conditions
(as shown in (asFigure
shown7).inIn Figure 7). Incase,
the latter the latter case, two photovoltaic
two photovoltaic panels were panels were
partially partially
shaded, in
shaded, in which the first was subjected to2 300 W/m22 solar radiation, while the second was
shaded,
which infirst
the which wasthe first
subjectedwas tosubjected
300 W/m to 300
solarW/m solar
radiation, radiation,
while the while
second the second
was was
subjected
subjected to 2500 W/m22 solar radiation.
to 500 W/m
subjected to 500solar
W/mradiation.
solar radiation.

Figure 14. PV system setup.

Figure
Figure 14.
14. PV
PV system
system setup.
setup.
The MPPT algorithms were evaluated under two different operating conditions of
The
The MPPT algorithms
MPPT algorithmswerewereevaluated
evaluatedunder under two different operating conditions of
the PV system. Figure 15a shows the first testingtwo different
condition operating conditions
characteristic of the
curves (case 1)
the PV
PVthe system.
system. Figure
Figure 15a 15a
shows shows the first testing condition characteristic curves (case 1)
of PV module, Figure 15bthe first testing
shows condition characteristic
the characteristic curves for casecurves (case
2, and 1) of 15c
Figure the
of
PVthe PV module,
module, Figure 15b shows the characteristic curves for case 2, and Figure 15c
shows case 3.Figure 15b shows
A transparent the
sheet characteristic
covered parts ofcurves
the PV for case
array to 2, and Figure
simulate 15c
partial shows
shading
shows
case 3. case
A 3. A transparent
transparent sheet sheet covered
covered parts parts
of the of
PV the PV array
array to to simulate
simulate partial
partial shading
shading with
with different patterns in cases 1, 2, and 3.
with different
different patterns
patterns in cases
in cases 1, 2,3.and 3.
1, 2, and
120-
120-
[W][W]

60 [W]
Power

60 - 60 [W]
Power

60 -
30 [W]
PVPV

30 [W]

0-
0 15 30
0- (a)
0 15 30
40- (a)
40-
29 [W] 27 [W]
[W][W]

29 [W] 27 [W]
Power

22 [W]
20-
Power

22 [W]
20-
PVPV

0-
- -

- -

00- 10 20
0 (b)
10 20
- (b)
-
36.5 [W]
[W][W]

36.5 [W]
Power

25 -
Power

25 - 12.5[W]
PVPV

12.5[W]

0-
0 7.5 15
0-
0 7.5
(c) 15
PV Voltage
(c) [V]
PV Voltage [V]

Figure 15. PV characteristics: (a) case 1, (b) case 2, (c) case 3.

Energies
Energies 2023,
2023, 16, x577
FOR PEER REVIEW 1212ofof18
17

The
The experimental
experimental results
results presented
presented in
in this
this section
section show
show thethe power
power extracted
extracted from
from the
the
PV
PV array
array and
and the
the voltage
voltage and
and current
current at
at the
the array’s
array’s output
output terminals
terminals obtained
obtained for
for each
each
MPPT algorithm.
MPPT algorithm.
Figure 16
Figure 16 illustrates
illustrates the
the power
power extracted
extracted from
from the
the PV
PV array
array operating
operating under
under partial
partial
shading conditions using the P&O algorithm. As shown in Figure 15, the
shading conditions using the P&O algorithm. As shown in Figure 15, the characteristic characteristic
curve Ppv ××Vpv
curve Ppv Vpvhashasone
oneGMPP
GMPPand andtwo
twoLMPP,
LMPP,withwiththe
theGMPP
GMPPequal
equaltoto 60
60 W
W and
and the
the
LMPP equal
LMPP equal to
to3030W.
W. The
TheP&OP&Omethod
methodreached
reachedthe theGMPP
GMPPfirst
firstand
andmaintained
maintainedthis
thisvalue.
value.

16.4 [V]

0 [V]

3.6 [A]

0 [A]

59.5 [W]

0 [W]
0.0 s 5.0 s
Figure 16. Experimental results of P&O MPPT for case 1.
Figure 16. Experimental results of P&O MPPT for case 1.
Figure 17 shows the power extracted from the PV array for the SA-based MPPT
Figure GMPP
algorithm. 17 showswasthe power extracted
achieved, from the
and the method PV arraya for
produced the of
power SA-based
60 W. InMPPT al-
addition,
this MPPT method resulted in reduced power fluctuations in steady-state.
gorithm. GMPP was achieved, and the method produced a power of 60 W. In addition, The results
demonstrate
this that theresulted
MPPT method SA-P&O-based MPPT
in reduced algorithm
power can quickly
fluctuations convert to GMPP
in steady-state. when
The results
the PV arraythat
demonstrate operates under partial shading
the SA-P&O-based conditions.
MPPT algorithm canIn addition,
quickly the proposed
convert to GMPP MPPT
when
method
the consistently
PV array operatesshowed
under less power
partial fluctuation
shading and greater
conditions. efficiency
In addition, in trackingMPPT
the proposed when
compared
method to the traditional
consistently showed method (P&O).fluctuation and greater efficiency in tracking
less power
whenFigures
compared16 and 17 traditional
to the show that the P&O(P&O).
method method takes longer to achieve the MPPT and
suffers from steady-state ripples more than the SA-P&O. Figure 18 shows the PV voltage,
current, and power when the SA-P&O method was used under partial shading conditions
when the maximum local and global forces, respectively, were 22 W and 29 W. GMPP was
achieved with low power oscillation and short convergence time. Experimental results
 16.5 [V]

Energies 2023, 16, 577 13 of 17

0 [V]

involving power, voltage, and current of the PV array of the P&O method are shown in
Figure
Energies 2023, 16, x FOR PEER REVIEW 19. The results indicates that the MPPT algorithm based on the P&O way did not 13 of 18
3.6 [A]
reach the GMPP, becoming fixed at an LMPP of 22 W. Moreover, the MPPT algorithm
showed more significant power fluctuations and had the longest convergence time.

0 [A]
16.5 [V]
60 [W]

0 [V]

0 [W]
3.6 [A]
0.0 s 5.0 s

Figure 17. Experimental results of SA-P&O MPPT considering case 1.

0 [A]
Figures 16 and 17 show that the P&O method takes longer to achieve the MPPT and
60 [W]from steady-state ripples more than the SA-P&O. Figure 18 shows the PV voltage,
suffers
current, and power when the SA-P&O method was used under partial shading conditions
when the maximum local and global forces, respectively, were 22 W and 29 W. GMPP was
achieved with low power oscillation and short convergence time. Experimental results
involving
0 [W] power, voltage, and current of the PV array of the P&O method are shown in
Figure 19. The results indicates that the MPPT algorithm based on the P&O way did not
0.0 s 5.0 s
reach the GMPP, becoming fixed at an LMPP of 22 W. Moreover, the MPPT algorithm
showed
Figure 17. more significant
Experimental resultspower fluctuations
of SA-P&O and had
MPPT considering the
case 1. longest convergence time.
Figure 17. Experimental results of SA-P&O MPPT considering case 1.

Figures 16 and 17 show that the P&O method takes longer to achieve the MPPT and
suffers from steady-state ripples more than the SA-P&O. Figure 18 shows the PV voltage,
16.5 [V]
current,
13.0 [V] and power when the SA-P&O method was used under partial shading conditions
when the maximum local and global forces, respectively, were 22 W and 29 W. GMPP was
0 [V]
achieved with low power oscillation and short convergence time. Experimental results
involving power, voltage, (a)and current of the PV array of the P&O method are shown in
Figure 19. The results indicates that the MPPT algorithm based on the P&O way did not
3.6 [A]
reach the GMPP, becoming fixed at an LMPP of 22 W. Moreover, the MPPT algorithm
2.2 [A]
showed more significant power fluctuations and had the longest convergence time.
0 [A]

16.5 [V] (b)

13.0 [V]

0 [V]
60 [W]

28.9 [W] (a)

3.60 [A]
[W]
2.2 [A]0.0 s (c) 10.0 s

0 [A]
Figure 18. Experimental results of SA-P&O MPPT considering Case 2 (a) PV voltage, (b) PV current,
and (c) PV power.

(b)

60 [W]
 Energies 2023, 16, x FOR PEER REVIEW 14 of 18

Energies 2023, 16, 577 14 of 17

Figure 18. Experimental results of SA-P&O MPPT considering Case 2 (a) PV voltage, (b) PV current,
and (c) PV power.

15 [V]
9 [V]
0 [V]

(a)

3.2 [A]
2.7 [A]

0 [A]

(b)

59.5 [W]

21.5 [W]
0 [W]
0.0 s (c) 10.0 s
Figure 19. Experimental results of P&O MPPT considering Case 2. (a) PV voltage, (b) PV current,
Figure
and (c) 19. Experimental results of P&O MPPT considering Case 2. (a) PV voltage, (b) PV current,
PV power.
and (c) PV power.
In case 3, the PV array was exposed to nominal irradiance, and then a part of the
In case 3,
PV was covered bythe
thePV array wassheet.
transparent exposed
Thetomethod
nominaltracked
irradiance, and then
the global MPPa part of the
at 36.5 W PV
without fixing on a local MPP, as shown in Figure 20. However, the P&O process fixed onwith-
was covered by the transparent sheet. The method tracked the global MPP at 36.5 W
localout
MPPfixing on a local
tracking MPP,
at 12.5 as shown
W and in find
failed to Figure
the20. However,
global theshown
MPP, as P&O process
in Figurefixed
21. on local
MPP tracking atthe
By comparing 12.5performance
W and failedoftoP&O
find the
andglobal MPP,
SA-P&O as shown
methods, it in Figure
was found 21.that
the tracking time was reduced by 18.5% when the latter method was used, and the most
important advantage was avoiding the LMPP.
Energies 2023, 16, 577 15 of 17
Energies
Energies 2023,
2023, 16,
16,xx FOR
FOR PEER
PEER REVIEW
REVIEW 15
15 of
of 18
18

16.5
16.5 [V]
[V]
13.5
13.5 [V]
[V]

00 [V]
[V]
(a)
(a)
3.6
3.6 [A]
[A]
2.7
2.7 [A]
[A]

00 [A]
[A]

(b)
(b)

60
60 [W]
[W]
36.5
36.5 [W]
[W]

00 [W]
[W]
0.0
0.0 ss (c)
(c) 10.0
10.0 ss

Experimental results
Figure 20.Experimental
Figure
of SA-P&O MPPT considering Case 3. (a)voltage,
PV voltage, (b) PV current,
Figure20.20. Experimentalresults
resultsof
ofSA-P&O
SA-P&OMPPT
MPPTconsidering
consideringCase
Case3.
3.(a)
(a)PV
PV voltage,(b)
(b)PV
PVcurrent,
current,
and
and (c)
and (c) PV
(c) PV power.
PV power.
power.

15
15 [V]
[V]
99 [V]
[V]
00 [V]
[V]

(a)
(a)

3.2
3.2 [A]
[A]
2.8
2.8 [A]
[A]

00 [A]
[A]

(b)
(b)

59.6
59.6 [W]
[W]

12.5
12.5 [W]
[W]
00 [W]
[W]
0.0
0.0 ss (c)
(c) 10.0
10.0 ss

Figure 21. Experimental results of P&O MPPT considering Case 3. (a) PV voltage, (b) PV current,
and (c) PV power.
Energies 2023, 16, 577 16 of 17

5. Conclusions
We present the results of implementing an SA-P&O-based MPPT algorithm for extract-
ing the maximum power from photovoltaic arrays. The efficiency of the MPPT algorithm
was validated by comparison with the MPPT P&O algorithm. From the results obtained
using computational simulation, it was demonstrated that the SA-P&O-based MPPT al-
gorithm was able to quickly converge to GMPP when the photovoltaic arrangement was
operating in STC, and when it is subjected to partial shading conditions. In addition, use of
the SA-P&O-based MPPT algorithm resulted in reduced power oscillations in steady-state
and greater tracking efficiency in the search of the GMPP.
Therefore, the SA-P&O-based MPPT method has an excellent performance in dealing
with local and global maximums in PV systems subjected to partial shading, maximizing
the system’s overall efficiency. Since our approach simulated conditions of real irradiation,
our results indicate excellent applicability of the technique to any PV system for delivering
maximum power output regardless of irradiation conditions and shading. As demonstrated
quantitatively, the proposed optimization method allows attainment of the best possible
power, which can guarantee maximum energy production from an entire PV system, with
the shortest financial return time for the investor.
Our results show that the SA-P&O method followed the GMPP efficiently in all STC
and PSC operating conditions and reduced transient periods, which increased the generated
energy and reduced the convergence time under all operating conditions. In addition to the
accurate converge to GMPP, SA-P&O requires fewer parameters than other metaheuristics
methods. The disadvantages of the proposed method are oscillations around the MPP, and
the complex computations required.

Author Contributions: Conceptualization, A.G.A.-K., I.I.E.-S. and A.R.; methodology, A.G.A.-K.,

I.I.E.-S. and A.R.; software, A.G.A.-K. and I.I.E.-S.; validation, A.G.A.-K., I.I.E.-S., and S.M.; formal
analysis, A.G.A.-K., I.I.E.-S., A.R. and S.M.; investigation, A.G.A.-K., I.I.E.-S. and A.R.; resources,
A.G.A.-K., I.I.E.-S. and A.R.; data curation, A.G.A.-K., I.I.E.-S. and A.R.; writing—original draft prepa-
ration, S.M., A.G.A.-K., I.I.E.-S. and A.R.; writing—review and editing, S.M. and A.R. Visualization,
A.G.A.-K. Supervision, A.G.A.-K., I.I.E.-S. and A.R. All authors have read and agreed to the published
version of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: Not applicable.
Acknowledgments: The authors would like to express their gratitude for international collaboration
between the UK, UAE and Egypt.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Almutairi, A.; Abo-Khalil, A.G.; Sayed, K.; Albagami, N. MPPT for a PV Grid-Connected System to Improve Efficiency under
Partial Shading Conditions. Sustainability 2020, 12, 10310. [CrossRef]
2. Brito, M.A.G.; Galotto, L.; Sampaio, L.P.; de Azevedo Melo, G.; Canesin, C.A. Evaluation of the main MPPT techniques for
photovoltaic applications. IEEE Trans. Ind. Electron. 2012, 60, 1156–1167. [CrossRef]
3. Premkumar, M.; Sumithira, T.R.; Sowmya, R. Implementation of Solar PV Based Microconverter with Optimal Mppt Control.
J. Electr. Eng. 2018, 18, 1–12.
4. Abo-Khalil, A.G.; Lee, D.C.; Seok, J.K.; Choi, J.W.; Kim, H.K. Maximum Power Point Tracking for Photovoltaic System Using
Fuzzy Logic Controller. In Proceedings of the KIPE Conference, The Korean Institute of Power Electronics, Hyundai Seongwoo
Resort, Gangwon, Republic of Korea, 14–17 June 2003; Volume 2, pp. 503–507.
5. Yu, B.G.; Abo-Khalil, A.G.; Matsui, M.; Yu, G. Sensorless Fuzzy Logic Controller for Maximum Power Point Tracking of Grid-
Connected PV system. In Proceedings of the International Conference on Electrical Machines and Systems ICEMS, Tokyo, Japan,
15–18 November 2009.
6. Jately, V.; Arora, S. Performance Investigation of Hill-Climbing MPPT Techniques for PV Systems Under Rapidly Changing Envi-
ronment. In Intelligent Communication, Control and Devices. Advances in Intelligent Systems and Computing; Singh, R., Choudhury, S.,
Gehlot, A., Eds.; Springer: Singapore, 2018; p. 624. [CrossRef]
Energies 2023, 16, 577 17 of 17

7. Jately, V.; Arora, S. An efficient hill-climbing technique for peak power tracking of photovoltaic systems. In Proceedings of the
2016 IEEE 7th Power India International Conference (PIICON), Bikaner, India, 25–27 November 2016; pp. 1–5. [CrossRef]
8. Veerachary, M.; Senjyu, T.; Uezato, K. Neural-network-based maximum-power-point tracking of coupled-inductor interleaved-
boost converter- supplied PV system using fuzzy controller. IEEE Trans. Ind. Electron. 2003, 50, 749–758. [CrossRef]
9. Elgendy, M.A.; Zahawi, B.; Atkinson, D.J. Assessment of perturb and observe MPPT algorithm implementation techniques for PV
pumping applications. IEEE Trans. Sustain. Energy 2012, 3, 21–33. [CrossRef]
10. Masood, B.; Siddique, M.S.; Asif, R.M.; Zia-ul-Haq, M. Maximum power point tracking using hybrid perturb & observe and
incremental conductance techniques. In Proceedings of the 2014 4th International Conference on Engineering Technology and
Technopreneuship (ICE2T), Kuala Lumpur, Malaysia, 27–29 August 2014; pp. 354–359.
11. Ahmed, J.; Salam, Z. An enhanced adaptive P&O MPPT for fast and efficient tracking under varying environmental condition.
IEEE Trans. Sustain. Energy 2018, 9, 1487–1496.
12. Hohm, D.P.; Ropp, M.E. Comparative study of maximum power point tracking algorithms. Prog. Photovolt. Res. Appl. 2003, 11,
47–62. [CrossRef]
13. Jately, V.; Azzopardi, B.; Joshi, J.; Venkateswaran, V.; Sharma, A.; Arora, S. Experimental Analysis of hill-climbing MPPT
algorithms under low irradiance levels. Renew. Sustain. Energy Rev. 2021, 150, 111467. [CrossRef]
14. Putri, R.I.; Wibowo, S.; Rifa’i, M. Maximum power point tracking for photovoltaic using incremental conductance method.
Energy Procedia 2015, 68, 22–30. [CrossRef]
15. Hameed, W.I.; Sawadi, B.A.; Muayed, A. Voltage Tracking Control of DC-DC Boost Converter Using Fuzzy Neural Network.
Int. J. Power Electron. Drive Syst. 2018, 9, 1657. [CrossRef]
16. Hussein, K.H.; Muta, I.; Hoshino, T.; Osakada, M. Maximum Photovoltaic Power Tracking: An Algorithm for Rapidly Changing
Atmospheric Conditions. IEE Proc. Gener. Transm. Distrib. 1995, 142, 59–64. [CrossRef]
17. Abo-Khalil, A.G.; Yu, G.Y. Current Estimation-based Maximum Power Point Tracker of Grid Connected PV. In Proceedings of the
10th International Conference on Power Electronics and Drives Systems (PEDS), Kitakyushu, Japan, 22–25 April 2013.
18. Kottas, T.L.; Boutalis, Y.S.; Karlis, A.D. New maximum power point tracker for PV arrays using fuzzy controller in close
cooperation with fuzzy cognitive networks. IEEE Trans. Energy Convers. 2006, 21, 793–803. [CrossRef]
19. Abo-Khalil, A.G.; Lee, D.C.; Choi, J.W.; Kim, S.G. Maximum Power Point Tracking Controller Connecting PV System to Grid.
Korean Inst. Power Electron. J. 2006, 6, 226–234.
20. Yu, G.Y.; Abo-Khalil, A.G.; Matsui, M.; Yu, G.J. Support Vector Regression Based Maximum Power Point Tracking for PV
Grid-Connected System. In Proceedings of the 2009 34th IEEE Photovoltaic Specialists Conference (PVSC), Philadelphia, PA,
USA, 7–12 June 2009.
21. Eltamaly, A.M.; Al-Saud, M.S.; Abo-Khalil, A.G. Photovoltaic maximum power point tracking under dynamic partial shading
changes by novel adaptive particle swarm optimization strategy. Trans. Inst. Meas. Control. August 2019, 42, 104–115. [CrossRef]
22. Pilakkat, D.; Kanthalakshmi, S. An improved P&O algorithm integrated with artificial bee colony for photovoltaic systems under
partial shading conditions. Sol. Energy 2019, 178, 37–47.
23. Eltamaly, A.M.; Al-Saud, M.S.; Abo-Khalil, A.G. A Novel Bat Algorithm Strategy for Maximum Power Point Tracker of
Photovoltaic Energy Systems Under Dynamic Partial Shading. IEEE Access 2020, 8, 10048–11006. [CrossRef]
24. Harrag, A.; Messalti, S. Variable step size modified P&O MPPT algorithm using GA-based hybrid offline/online PID controller.
Renew. Sustain. Energy Rev. 2015, 49, 1247–1260.
25. Seyedmahmoudian, M.; Rahmani, R.; Mekhilef, S.; Oo, A.M.; Stojcevski, A.; Soon, T.K.; Ghandhari, A.S. Simulation and hardware
implementation of new maximum power point tracking technique for partially shaded PV system using hybrid DEPSO method.
IEEE Trans. Sustain. Energy 2015, 6, 850–862. [CrossRef]
26. Chao, R.M.; Nasirudin, A.; Wang, I.K.; Chen, P.L. Multicore PSO operation for maximum power point tracking of a distributed
photovoltaic system under partially shading condition. Int. J. Photoenergy 2016, 2016, 9754514. [CrossRef]
27. Elhaddad, Y. Combined Simulated Annealing and Genetic Algorithm to Solve Optimization Problems. World Acad. Sci.
Eng. Technol. 2012, 6, 1047–1049.
28. Eltamaly, A.M.; Al-Saud, M.S.; Abo-Khalil, A.G. A novel scanning bat algorithm strategy for maximum power point tracker of
partially shaded photovoltaic energy systems. Ain Shams Eng. J. 2020, 11, 1093–1103. [CrossRef]
29. Sharma, A.; Sharma, A.; Jately, V.; Averbukh, M.; Rajput, S.; Azzopardi, B. A Novel TSA-PSO Based Hybrid Algorithm for GMPP
Tracking under Partial Shading Conditions. Energies 2022, 15, 3164. [CrossRef]
30. Jately, V.; Bhattacharya, S.; Azzopardi, B.; Montgareuil, A.; Joshi, J.; Arora, S. Voltage and Current Reference Based MPPT Under
Rapidly Changing Irradiance and Load Resistance. IEEE Trans. Energy Convers. 2021, 36, 2297–2309. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.