A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques

Contents lists available at ScienceDirect
Solar Energy
journal homepage: www.elsevier.com/locate/solener
Review
A novel evaluation index for the photovoltaic maximum power point tracker
techniques
Ali M. Eltamalya,b
, Hassan M.H. Farhc,d,⁎
, Mohd F. Othmanc
a
Electrical Engineering Department, Mansoura University, Mansoura, Egypt
b
Sustainable Energy Technologies Center, King Saud University, Riyadh 11421, Saudi Arabia
c
Malaysia-Japan International Institute of Technology, Universiti Teknologi Malaysia, KL 54100, Malaysia
d
Eectrical Engineering Department, College of Engineering, King Saud University, Riyadh 11421, Saudi Arabia
A R T I C L E I N F O
Keywords:
Partial shading
Global peak
Evaluation index
MPPT techniques
Artificial intelligence
Bio-Inspired
A B S T R A C T
The partially shaded photovoltaic (PSPV) condition reduces the generated power and contributes in hot spot
problem. PSPV generates one global peak (GP) and many local peaks (LP) in power versus voltage curve. In
recent years, numerous research papers have been focused on highly efficient maximum power point tracking
(MPPT) techniques to track the GP and alleviate the partial shading effects. This paper provides a comparative
and comprehensive review of the 17 most famous and efficient MPPT techniques. These famous and efficient
MPPT techniques are divided into three groups; conventional, soft computing (Artificial Intelligence and Bio-
Inspired) and hybrid MPPT techniques. Technical and economical comparisons of these 17 MPPT techniques
based on 17 evaluation parameters are then achieved. The findings obtained have not yet been discovered yet
before where this is the first time the 17 most famous and efficient MPPT techniques are ranked using a novel
evaluation index with a total evaluation from 40 points based on the 8 most important key issues. These issues
are tracking speed, convergence speed, complexity, hardware implementation, initial parameters required,
performance without PS, performance with PS, and efficiency. Finally, merits, demerits, technical and eco-
nomical comparisons of all MPPT techniques are also introduced, discussed, and assessed.
1. Introduction
Solar photovoltaic (PV) energy system is considered as one of the
most promising technologies of renewable generation systems because
it is clean, abundant, noise free, and friendly to the environment
compared to conventional energy sources such as natural gas, or any
other fossil fuels. To maximize the energy captured from a PV system,
maximum power point tracking (MPPT) should be used. Tracking the
maximum power from a PV system is considered a hot research area, as
it can improve the system’s efficiency, reliability, power quality and
flexibility (Arunkumari and Indragandhi, 2017; Babu et al., 2015).
The unique peak under uniform condition (without partial shading)
can be tracked efficiently and accurately using conventional MPPT
techniques. Whereas, multiple peaks; one global peak (GP) and many
local peaks (LPs) are generated under partial shading conditions (PSCs).
Therefore, highly efficient and modern MPPT techniques based on
evolutionary, heuristic and metaheuristic techniques should be carried
out to track the GP instead of LPs, after failure of conventional tech-
niques to track the GP in some cases of PSCs. In recent years, numerous
review papers (Ramli et al., 2017; Ishaque and Salam, 2013; Liu et al.,
2016; Kamarzaman and Tan, 2014; Kandemir et al., 2017; Mohapatra
et al., 2017; Salam et al., 2013; Seyedmahmoudian et al., 2016; Karami
et al., 2017; Ram et al., 2017) are focused on the MPPT techniques
under uniform condition and PSCs including the idea of operation, GP
tracking efficiency, and classifications. The PV-MPPT techniques have
been classified into conventional, soft computing, and hybrid techni-
ques in many review researches depending on the MPPT technology
(Ramli et al., 2017; Ishaque and Salam, 2013; Liu et al., 2016;
Kamarzaman and Tan, 2014; Kandemir et al., 2017). Conventional
MPPT techniques include Perturb and Observe (P&O), Incremental
Conductance (IC), Hill Climbing (HC), and Constant Voltage (CV)
technique, I-V Curve Scanning-Tracking, Fibonacci Searching, global
MPPT Segmentation Searching, extremum seeking control, and etc.
Whereas, soft computing techniques or sometimes called modern
MPPT use evolutionary, heuristic, and metaheuristic algorithms based
on Artificial Intelligence (AI) or Bio-Inspired (BI) such as Fuzzy Logic
Control (FLC), Artificial Neural Network (ANN), Differential Evolution
(DE), Genetic Algorithm (GA), Particle Swarm Optimization (PSO),
https://doi.org/10.1016/j.solener.2018.09.060
Received 8 May 2018; Received in revised form 13 September 2018; Accepted 20 September 2018
⁎
Corresponding author at: Eectrical Engineering Department, College of Engineering, King Saud University, Riyadh 11421, Saudi Arabia.
E-mail address: hfarh1@ksu.edu.sa (H.M.H. Farh).
Solar Energy 174 (2018) 940–956
0038-092X/ © 2018 Elsevier Ltd. All rights reserved.
T

Simulated Annealing (SA), Tabu Search (TS), Cuckoo Search
Optimization (CSO), Teaching Learning Based Optimization (TLBO),
Firefly Algorithm (FA), Flower Pollination Algorithm (FPA), Ant Colony
Optimization (ACO), Ant Bee Colony (ABC), Grey Wolf Optimization
(GWO), Fibonacci Line Search (FLS), Improved Curve Tracer, Improved
Extremum-Seeking, Simulated Annealing, Variable Step Newton-
Raphson, Variable Step Size P&O, Optimal P&O based on Least Square
Support Vector Machines, Dynamic Population Size DE, Chaotic Search
and etc. Finally, hybrid techniques that can effectively track the GP are
PSO-P&O, GWO-P&O (BI-Conventional), DE-PSO (AI-BI), FLC-GA or
ANN-GA (AI-AI), GA-P&O (AI- Conventional), and etc. (Ramli et al.,
2017; Ishaque and Salam, 2013; Liu et al., 2016; Kamarzaman and Tan,
2014; Kandemir et al., 2017; Mohapatra et al., 2017). Due to the merits
from MPPT on the performance and cost of energy generated from PV
system, a huge number of researches have been introduced to the sci-
entific community which have proved their importance. A database
collection of the recent researches in the area of MPPT for the most
important and prestigious two publication houses like Elsevier and IEEE
shown the fast grouping in this research topic in the last 8 years as
summarised in Table 1 and Fig. 1. Table 1 shows the number of the PV
MPPT publications (articles, books and others) from 2010 to 2017 for
Elsevier and IEEE. It concludes from Fig. 1 that a sharp growth of the
total publications number of the PV MPPT between 2010 and 2017 is
evidence to prove the importance of this topic. Therefore, it reflects the
worldwide focus on the maximum power extraction from the PV energy
systems. Although tracking the MPP can improve the performance of
PSPV system, the PV system topologies like the bypass diodes, PV
system architectures, PV array configuration and PV array re-
configuration can mitigate and alleviate the partial shading (PS) effects
on the PV array itself and the power extracted from it. The most popular
MPPT techniques, PV array topologies, architectures and configurations
have been discussed in (Kandemir et al., 2017). The PV array topologies
are divided into multi or two stage (dc-dc and dc-ac) PV system, single
stage, and Module Integrated Converter (MIC). Whereas PV system
architectures are classified as centralized, string connected, series MIC,
parallel MIC and Sub-MIC. In addition, the most popular PV array
configurations (Series-Parallel; SP, Total Cross Tie; TCT and Bridge
Linked; BL) are introduced to remove the local peaks (Kandemir et al.,
2017). On the other hand, Mohapatra et al. discussed various converter
topologies (Cascaded H-Bridge, MIC; to track the MPP directly and
accurately, multi-level inverter with PV groups; for independent MPPT
control, shunt-series compensation and variable interleaved dc-dc
converter). In addition, they presented modern and hybrid MPPT
techniques with new PV modeling approach under PS (Fast power
peaks estimator during PSPV systems to track the MPP and sub-module
integrated converter to reduce power loss) (Mohapatra et al., 2017).
A comprehensive review of the soft computing MPPT techniques
(ANN, non-linear predictor, chaotic search, FLC, PSO, ACO, GA, DE and
Bayesian network) are discussed to evaluate their performance based
on 5 evaluation parameters which are PV array dependency, con-
vergence speed, ability to handle PSC, complexity and hardware im-
plementation (Salam et al., 2013). AI based MPPT techniques (ANN,
FLC, PSO, ACO, GA, DE, CSO, GWO and FA) for mitigating the PS ef-
fects based on 10 evaluation parameters (convergence speed, system
independency, performance with PS, performance without PS, effi-
ciency, complexity, hardware implementation, periodic tuning, de-
pendency of the initial, oscillation around MPP) have been assessed and
focused on their background, theory, performance, limitations and
significant features (Seyedmahmoudian et al., 2016). Whereas, Karami
et al. (2017) presented a general review and comparisons of 40 old and
recent MPPT techniques based on 5 evaluation parameters which are
analog or digital, sensors used, speed, stability and periodic tuning. The
40 MPPT techniques are classified according to their MPP tracking
technology to five categories; predefined fixed parameters, measure-
ment and comparison with a pre-known MPP, trial and error or cal-
culation and observe, mathematical calculation, and intelligent pre-
diction (Karami et al., 2017). The previous studies are supported by
Ram’s study in 2017 (Ram et al., 2017) which presented the state of the
art review on various MPPT techniques covering conventional (P&O, IC,
HC and Global MPPT) and recent soft computing (FLC, ANN, GA, PSO,
CSO, ACO, ABC, FA, Random search and non-linear) techniques. A
comprehensive comparison between these conventional and soft com-
puting techniques based on 4 evaluation parameters which are tracking
speed, complexity dynamic tracking under PSCs and hardware im-
plementation is then evaluated (Ram et al., 2017).
On the basis of the comprehensive literature review, there are nu-
merous numbers of research papers focused on modern and efficient
MPPT techniques to track the GP under PSCs. For this purpose, this
paper not only provides a comparative and comprehensive review of 17
famous and efficient MPPT techniques but, it also evaluates and ranks
them according to a novel evaluation index. The 17 most famous and
efficient MPPT techniques have been covered and classified into three
groups which are conventional, soft computing (AI and BI), and hybrid
Table 1
Number of the photovoltaic MPPT publications from 2010 to 2017.
Article type Years
2010 2011 2012 2013 2014 2015 2016 2017
Elsevier Articles 1257 1800 2101 2727 3471 3907 4620 5795
Book chapters 97 82 127 214 194 215 221 297
Others 91 139 230 214 167 155 194 234
All 1445 2021 2458 3155 3832 4277 5035 6326
IEEE Articles 369 482 576 692 857 859 1084 735
Books 0 1 0 0 1 1 2 1
Others 0 0 2 0 0 0 0 0
All 369 483 578 692 858 860 1086 736
Total publications 1814 2504 3036 3847 4690 5137 6121 7062
0
1000
2000
3000
4000
5000
6000
7000
8000
2010 2011 2012 2013 2014 2015 2016 2017
1445
2021 2458
3155
3832 4277
5035
6326
369
483
578
692
858
860
1086
736
All Elsevier publications no. All IEEE publications no.
Fig. 1. Elsevier and IEEE MPPT publications number from 2010 to 2017.
A.M. Eltamaly et al. Solar Energy 174 (2018) 940–956
941

MPPT techniques as shown in Fig. 2. On the other hand, previous re-
searches did not introduce an evaluation index to rank these MPPT
techniques to help researchers, scientists, and industrial sector to pick
up the most effective technique easily. For this matter, a novel eva-
luation index for the PV MPPT techniques is introduced in this paper.
Technical and economical evaluation for 17 MPPT techniques are in-
troduced based on 17 evaluation parameters in addition to the total
evaluation with 40 points for these MPPT techniques are achieved
based on the 8 most important evaluation parameters. Studying and
analyzing numerous numbers of researches, comparative and compre-
hensive review papers in the area of MPPT helped us to introduce a new
evaluation index to evaluate and rank the 17 most important MPPT
techniques with and without PSCs.
2. Description of the partial shading photovoltaic system
The PV-modules should be connected in parallel and series to in-
crease the current and voltage, respectively to be suitable for the load
requirements. Partial shading condition (PSC) happens when one or
more PV-modules in the PV array are exposed to different radiation.
When PS occurs, shaded PV-modules will face a current higher than the
generated current and it will act as a load for the other PV-cells. Due to
the increased current flow in the shaded PV-modules which is higher
than its generated current, the voltage will become negative across this
PV-modules and it can be higher than the rated voltage of these mod-
ules which can destroy it by forming hot spot problem (Femia et al.,
2005). Fig. 3 shows the general scheme of the PSPV system. The PV
arrays are interconnected to the utility grid through a dc-dc converter
and three phase inverter. Under uniform condition, a unique MPP will
be generated that can be tracked easily and efficiently using conven-
tional, soft computing, or hybrid MPPT techniques. On the other hand,
under PSC, different radiation on each PV array generates different
power from one PV array to another and multiple peaks (one GP and
many LPs) will be generated due to bypass diodes used for protecting
the PV arrays from the hot spot points and thermal breakdown. The
maximum power available from the PSPV system is equal to the GP.
Three different shading patterns/cases occur under PSCs (Case 1; GP at
the beginning, Case 2; GP at the middle, Case 3; GP at the end) as shown
in Fig. 4. Numerous modern and efficient MPPT techniques are carried
out to tack the GP instead of the LP. Each MPPT technique has its own
merits and demerits in addition to the input depending on the MPPT
technique used. It may be irradiance, temperature, PV voltage and
current whereas the output is the optimal duty ratio of the dc-dc con-
verter as shown in Fig. 3. The dc-dc converters used to track the GP may
be buck, boost, buck-boost, flyback, SEPIC converters, whereas, the
most famous one is the boost converter because many PV applications
need to boost the output voltage to a higher value to be suitable for
loads.
3. Maximum power point tracking techniques
Under uniform condition, the P-V characteristic contains a unique
MPP for each weather condition (radiation and temperature) as shown
in Fig. 5. This unique MPP can be tracked efficiently and accurately
through conventional techniques and there is no need for more so-
phisticated techniques like soft computing or hybrid MPPT techniques.
On the other hand, the P-V characteristic contains multiple peaks (one
GP and many LPs) under non-uniform or PSC as shown previously in
Fig. 4. Therefore, the AI, BI or hybrid techniques based MPPT will be
effective, efficient, accurate and reliable in tracking the GP because
most of the conventional techniques may be stuck at one of the LP. In
this paper, MPPT techniques are divided into two categories based on
the suitability for PSCs which are MPPT techniques with and without
PS. Based on previous comparative studies of MPPT techniques, merits
and demerits of the famous, effective and efficient 17 MPPT techniques
with and without PS will be summarized in the next section. Also,
technical and economical comparisons of these 17 MPPT techniques
will be introduced based on 17 evaluation parameters collected from
previous comparative and comprehensive review studies till the end of
2017. A total evaluation of all these MPPT techniques with and without
MPPT Techniques
Conventional Soft Computing Hybrid
HC AI BI
P&O
IC
CV
DE
GA
FLC
ANN
ANFIS
PSO
CSO
TLBO
ACO
ABC
FA
FPA
GWO
ANN-IC
ANN-P&O
Conventional/Soft
Soft/Conventional
Soft/Soft
P&O-GA
ANN-FLC
FLC-GA
GA-ANN
Fig. 2. Classification of PV MPPT techniques.
=
=
=
dc-dc converter Three phase inverter
VPV IPV
Utility Grid
PV
array 1
PV
array 2
PV
array 3
VSC
control
P&O
IC
HC
CV
DE
GA
FLC
ANN
PSO
TLBO
CSO
ACO
ANFIS
ABC
FA
FPA
GWO
Conventional MPPT techniques
AI techniques BI techniques
Irradiance Temperature
Fig. 3. General scheme of the PSPV system.
Case 1
Case 3
Case 2
15
10
5
0
PV
power,
W
PV Voltage, V
1 2 3 4 5 6
0
Fig. 4. Three different GP cases under PSCs.
942

PS is introduced based on the 8 most important evaluation parameters
with 40 points. Each evaluation parameter has five points as a weight
and some evaluation parameters have positive trend while others have
negative trend from one to five for very low, low, medium, high and
very high. Therefore, the 17 selected MPPT techniques with and
without PS are evaluated and ranked.
3.1. Conventional MPPT techniques
Numerous conventional MPPT techniques have been used to track
the unique MPP under uniform condition (without PS). The most fa-
mous conventional techniques are P&O, IC, HC, and CV. The basic idea
of operation, literature review, merits and demerits in addition to
comparisons of these conventional techniques have been introduced in
the following subsections. The comparisons made based on previous
comparative and comprehensive review studies untill recently for 17
evaluation parameters are as follows:
3.1.1. Perturb and observe MPPT technique
The Perturb and Observe (P&O) based MPPT uses the operating
voltage perturbation of the array and observe the output power varia-
tion. If the power increases with the last voltage increment, then, the
next perturbation must be kept in the same direction. On the other
hand, if the voltage increment reduces the power, the next perturbation
must be reversed. The maximum power is achieved when dp/dV = 0
(Khadidja and Mountassar, 2017; Houssamo et al., 2010; Salas et al.,
2006; Sharma and Katti, 2017; Noman et al., 2012).
P&O is widely used due to its simple implementation. However,
oscillations around MPP at steady state during rapid change of radia-
tion represent the main shortcomings of this technique. Numerous re-
search papers are proposed to overcome and face this problem. For
example, Femia et al. (2005) optimized the P&O performance through
the customization of the two main parameters of P&O which are duty
cycle perturbation (Δd) and sampling time (Ta). Reducing Δd prevent
oscillation around MPP while increasing Ta avoid instability of MPP
and track quick varying MPP during rapid change of radiation (Femia
et al., 2005). This is supported by the study of Pandey et al. (2008)
which proposed a variable-step-size Δt P&O algorithm for drift
avoidance at steady state and fast tracking of MPP (Pandey et al.,
2008). Also, Abdelsalam et al. (2011) proposed adaptive P&O MPPT
technique to achieve adaptive tracking, no steady-state oscillations
around the MPP, and lastly, no need for predefined system-dependent
constants (Abdelsalam et al., 2011). A recent study by Ahmed and
Salam (2015) proposed an improved P&O for higher efficiency by re-
ducing the steady state oscillation and eliminating the possibility of the
algorithm to lose its tracking direction through a dynamic perturbation
step-size. The improved P&O is compared to the conventional and
adaptive P&O and it was found that, the MPPT efficiency increased by
2% compared to the other two techniques (Ahmed and Salam, 2015).
The development and experimental comparisons of P&O and IC algo-
rithms have been carried out in (Houssamo et al., 2010). The findings
proved that optimized P&O can have mostly the same efficiency as IC
and outperformed other techniques by its simple implementation
(Houssamo et al., 2010). The above finding is consistent with the study
by Ghassami et al. (2013). They revealed that, the modified P&O and IC
techniques are efficient and accurate to extract the maximum power
under a rapid variation of environmental conditions. It uses the I-V
curve to discriminate between rapid change of radiation (MPP varies)
and move the operating point to the MPP in fixed radiation (Ghassami
et al., 2013).
Finally, the merits and demerits of P&O based MPPT technique is
introduced in Table 2. On the other hand, previous comparative studies
of P&O with other conventional MPPT techniques have been summar-
ized in Table 3. The majority of comparative studies findings proved
that P&O performed less than IC but outperformed the other conven-
tional techniques (Ishaque et al., 2014; Jeddi and Ouni, 2014; Tofoli
et al., 2015; Cavalcanti et al., 2007; Gupta et al., 2016; Zainudin and
Mekhilef, 2010). In addition, some authors proved that some mod-
ifications and improvements on P&O put it in the same level of per-
formance with IC (Houssamo et al., 2010; Hohm and Ropp, 2003;
Faranda and Leva, 2008).
3.1.2. Incremental Conductance MPPT technique
IC technique depends on the slope of the P-V array characteristics
where MPP is obtained when dP/dV = 0 as shown in the following
equations (Ishaque et al., 2014):
d V I
dV
I V
dI
dV
( , )
0
PV PV
PV
PV PV
PV
PV
= + =
(1)
dI
dV
I
V
PV
PV
PV
PV
=
(2)
The current variation, dIPV and the voltage variation, dVPV are ap-
proximated to ΔVPV and ΔIPV as follows:
dV V V t V t
( ) ( )
PV PV PV PV
2 1
= (3)
dI I I t I t
( ) ( )
PV PV PV PV
2 1
= (4)
When
dI
dV
I
V
PV
PV
PV
PV
= is satisfied, then the MPP is reached, and the
operating point is exactly equal to MPP. If the operating point dpPV/
dVPV is greater than zero (
dI
dV
I
V
PV
PV
PV
PV
), then the operating point is on the
left of MPP of the P-V curve. On the other hand, if it is less than zero
(
dI
dV
I
V
PV
PV
PV
PV
), the operating point is on the right of the of the P-V curve.
Table 3 shows the previous comparative studies of the conventional
Table 2
Merits and demerits of P&O technique.
Ref. Merits Demerits
Jeddi and Ouni (2014), Tofoli et al. (2015), Gupta et al. (2016),
Danandeh and Mousavi (2017), Nabipour et al. (2017) and
Bendib et al. (2015)
• Simple in construction and
implementation
• Straightforward, accurate and high
performance without PS
• Online and independent on PV array
• Oscillations around MPP at steady state occurred
during sudden or fast change in atmospheric
conditions
• Controlling the perturbation size is difficult
0 1 2 3 4 5 6
0
5
10
15
20
Terminal Voltage of PV Module (V)
Generated
Power
(W)
Fig. 5. P-V characteristics under uniform conditions.
943

MPPT techniques which proved that IC is the most efficient and robust
conventional technique compared to other conventional ones (P&O,
HC, CV, OV, SC, TP, PC, TPE, TS and Fixed duty cycle) in terms of
steady state error, dynamic response and efficiency followed by P&O
technique (Zainudin and Mekhilef, 2010; Jeddi and Ouni, 2014; Tofoli
et al., 2015; Cavalcanti et al., 2007).
The finding is consistent with the findings of past studies by Gupta
et al. (2016) which revealed that, IC has superior performance as
compared to P&O and CV in terms of tracking efficiency, rise time, fall
time and dynamic response (Gupta et al., 2016). Similarly, Ishaque
et al. found that IC performance is slightly better than P&O and it is
very sensitive to its perturbation size, especially at low irradiance levels
(Ishaque et al., 2014). On the other hand, both Faranda and Hohm
concluded that P&O and IC have superior and similar performance in
addition to higher efficiency compared to other conventional techni-
ques (Hohm and Ropp, 2003; Faranda and Leva, 2008). Merits and
demerits of IC based MPPT are shown in Table 4.
3.1.3. Hill Climbing MPPT technique
The Hill Climbing (HC) technique is very simple in logic, im-
plementation and priori information is not required. The basic idea of
operation depends on using the converter duty cycle perturbation and
determines the variation of the power until the change in power be-
comes zero value to locate the MPP. Rapid fluctuations of solar radia-
tion may cause the algorithm to lose fast tracking of the MPP com-
pletely due to lack of fast response. Also, oscillations occur around MPP
at steady state during fast change in atmospheric conditions (Rezk and
Eltamaly, 2015; Shimizu et al., 2003; Koutroulis et al., 2001; Xiao and
Dunford, 2004; Veerachary et al., 2001). Merits and demerits of HC
based MPPT are shown in Table 5.
3.1.4. Constant voltage MPPT technique
Constant Voltage (CV) technique forces the PV array’s voltage to a
fixed value where the MPP voltage (VMPP) is approximated to 76% of
the PV array’s open circuit voltage (VOC) (Cavalcanti et al., 2007). The
shortcomings of this technique are that the VMPP is not always at 76% of
the VOC, it increases the steady state error hence reducing the efficiency.
The CV controller has some merits such as only one voltage sensor is
needed and the current sensor is not required (Faranda and Leva,
2008). Also, it is the easiest technique in implementation and has low
installation cost, but its efficiency is poor with respect to other active
MPPT techniques. The block diagram of CV controller is shown in Fig. 6
where VPV is only measured in order to provide the duty cycle of the dc-
dc converter by PI regulator to track the MPP (Gupta et al., 2016).
Table 4
Merits and demerits of IC technique.
Jeddi and Ouni (2014), Gupta et al. (2016), Danandeh and
Mousavi (2017), Nabipour et al. (2017) and Bendib et al.
(2015)
• Online, faster, more accurate, reliable and
efficient
• Variable perturbation size makes it more
adaptable to fast changing conditions
• Oscillation around MPP is less
• Response time is longer during atmospheric
conditions variation
• More expensive
• Speed and accuracy depend on increment size
hence oscillations may be occurred.
Table 3
Previous comparative studies of conventional MPPT techniques.
Ref. year MPPT Variable
control
dc-dc
converter
Application Simulation/
Experimental
Findings
Ishaque et al.
(2014)
2014 IC, P&O Duty cycle Buck-boost Standalone PV Simulation IC performance is slightly better than P&O and is very sensitive
to its perturbation size, especially at low irradiance
Faranda and
Leva (2008)
2008 IC, P&O, CV, OV*
,
SC*
and TP*
Duty cycle SEPIC*
Grid-connected Simulation P&O and IC have superior and similar performance in addition
to higher efficiency compared to other techniques
Jeddi and Ouni
(2014)
2014 IC, P&O, FOD*
Duty cycle Buck Standalone PV Simulation IC technique has the best MPPT error (best tracking) and high
efficiency followed by P&O and finally FOD
Tofoli et al.
(2015)
2015 IC, P&O, Fixed
duty cycle and CV
Duty cycle Buck Standalone PV Simulation IC has a good performance followed by P&O compared to other
techniques in terms of efficiency, tracking speed and steady-
state error. However, CV is simple and uses only one voltage
sensor, but efficiency is poor. Fixed duty cycle is not adequate
for high power PV system
Elgendy et al.
(2012)
2012 Two P&O (ΔVref &
ΔD)
Duty cycle
and Voltage
Buck Standalone PV
pumping systems
Both Theoretical and experimental comparison of the two P&O
implementation techniques (ΔVref & ΔD) for the suitable choice
of main parameters on the basis of system stability,
performance characteristics, and energy utilization
Hohm and Ropp
(2003)
2003 P&O, IC, PC, CV Duty cycle Buck Standalone PV Both IC and P&O performed well in terms of MPPT efficiency that
makes them favourable over the simpler CV and PC
Cavalcanti et al.
(2007)
2007 IC, CV, HC, P&O,
TPE*
, TS*
Duty cycle Boost Both Both IC is the most efficient and robust in terms of steady state error,
dynamic response and efficiency. TPE cannot track the true
MPP with sudden change of irradiance
Gupta et al.
(2016)
2016 (IC, CV, P&O) &
(ANFIS, FLC,
ANN)&(MP&O,
PI-FLC and N-FLC)
Duty cycle Boost Standalone PV Simulation IC has superior performance compared to P&O and CV in terms
of tracking efficiency, rise time, fall time and dynamic
response. ANFIS has better tracking efficiency than FLC and
ANN. Neural-FL has better efficiency than other conventional
and hybrid MPPT techniques
Zainudin and
Mekhilef
(2010)
2010 IC, P&O Duty cycle Buck, boost
and cuk
Standalone PV Simulation IC performs well and has a better output value than P&O
regardless of whether the dc-dc converter is buck or boost or
cuk converter
Houssamo et al.
(2010)
2010 P&O, IC Duty cycle Boost Standalone PV Both Optimized P&O can have almost the same efficiency as INC and
outperformed other techniques by its easiest implementation
*
OV: Open voltage; SC: Short current pulse; TP: Temperature Parametric; FOD: First order differential; PC: Parasitic capacitance; TPE: Tolerable Power Error; TS: Two
Stages of operation; in the first stage, variable large steps allow fast tracking when the PV voltage is far from the MPP voltage. The second stage with any technique
using fixed step can be used to track the MPP. *
SEPIC: Single Ended Primary Inductor Converter.
944

Merits and demerits of CV based MPPT technique is introduced in
Table 6.
3.2. Soft computing MPPT techniques
3.2.1. Artificial Intelligence based MPPT techniques
3.2.1.1. Fuzzy logic control technique. The objective of FLC is to track and
extract maximum power from the PV system for a given irradiance
(W/m2
) and temperature (°C). It does not require any technical
knowledge for the PV system, while its simplicity gives it an advantage
in tracking its MPP under fast varying atmospheric conditions (Ansari
et al., 2010; Azzouzi, 2012). The FLC has two inputs which are dP
dV
PV
PV
and
( )
dP
dV
PV
PV
i.e. (Err) and (ΔErr) which are determined from the PV output
power and the voltage (Fuzzification) as follows:
E
P k P k
V k V k
( ) ( 1)
( ) ( 1)
rr
PV PV
PV PV
=
(5)
E E k E k
( ) ( 1)
rr rr rr
= (6)
The output from FLC is the required change in the duty cycle of the
dc-dc converter (De-Fuzzification). The FLC block diagram is shown in
Fig. 7. The advantages of using FLC are it is being a universal control
algorithm, very simple, adaptive, fast tracking response, parameter
insensitivity and it can work properly even with an imprecise input
data. Also, FLC has better and efficient response in tracking the MPP,
especially in case of rapid changing atmospheric conditions
(Kamarzaman and Tan, 2014; Chekired et al., 2014; Mahamudul et al.,
2013; Messai et al., 2011). One of its drawbacks occurs in PSC where it
may stick around LP.
Fig. 8 shows the inputs and output membership functions and
Table 7 introduces the inputs and output fuzzy rules (Rezk and
Eltamaly, 2015). The variation step of Err and ΔErr may vary according
to the system. Once Err and ΔErr are calculated and transferred to the
logic variables based on membership functions, the FLC output, which
is typically duty ratio change, ΔD of the dc-dc boost converter is esti-
mated in rules as shown in Table 7.
Based on recent comparative studies of FLC based MPPT shown in
Table 8, it can be concluded that, FLC has faster convergence speed in
tracking the unique peak under uniform conditions compared to the
conventional MPPT techniques (Rezk and Eltamaly, 2015; Bendib et al.,
2014; El Khateb et al., 2013). Also, adaptive FLC has better perfor-
mance compared to the direct and indirect FLC based MPPT during
dynamic and steady state conditions regardless of the converter type
(Nabipour et al., 2017; Kwan and Wu, 2016; Guenounou et al., 2014).
In conclusion, FLC should be combined with a scanning and storing
algorithm or other AI techniques to track the GP with PS for achieving
fast and accurate convergence, high tracking efficiency and drift
avoidance (Boukenoui et al., 2016). Merits and demerits of FLC based
MPPT technique is shown in Table 9.
3.2.1.2. Artificial neural network technique. Artificial Neural Network
(ANN) represents one of the artificial intelligent MPPT techniques that
have the ability to solve nonlinear problems. Therefore, it can be
applied to track the GP over the LPs. ANN consists of three layers; input,
hidden, and output layers. The input layers are defined from the PV
array such as temperature, irradiance and Isc or Vo.c (Ishaque and Salam,
2013; Kamarzaman and Tan, 2014; Rai et al., 2011). ANN adjusts and
controls the duty cycle of the dc-dc converter (ANN output) to track the
GP.
Based on recent comparative studies of ANN based MPPT shown in
Table 10, it can be observed that ANN is efficient and accurate in
tracking the unique peak under uniform condition compared to con-
ventional techniques and mitigate their shortcomings related to
tracking speed and oscillations around MPP at steady state (Rai et al.,
2011; Messalti et al., 2017; Laudani et al., 2014; Mancilla-David et al.,
2014). On the other hand, ANN is preferred if combined with other
conventional or AI MPPT techniques to extract the GP instead of LP
from the PV array where, ANN is used to predict the GP region whereas
conventional or AI technique is used to track the GP. The reasons be-
hind these are irradiance sensors are relatively expensive or may not be
available, in addition to sufficient training needs a huge number of data
points that increases network complexity and is time consuming espe-
cially for PSC. Also, enlarged optimization scope for the size and hidden
layers number and retraining due to system aging is required as a result
of PV characteristics change. For example, Punitha et al. combined ANN
with IC track the GP efficiently compared to P&O and FLC based HC
(Punitha et al., 2013), while Jiang et al. combined ANN with P&O
where ANN is used to predict GP searching area and P&O to track the
GP. The findings revealed that, the proposed hybrid MPPT can track the
GP efficiently and accurately compared to P&O, Fibonacci search,
conventional PSO and DE (Jiang et al., 2015). Also, Loubna et al.
proved that ANN with a scanning and storing algorithm has better
performance than variable P&O with global scanning and IC based on
FLC (Bouselham et al., 2017). Finally, Karatepe et al proposed ANN
combined FLC with polar information controller to track the unique
peak under uniform condition and the GP under PSCs. The ANN is
Table 5
Merits and demerits of HC based MPPT technique.
Danandeh and Mousavi (2017), Nabipour et al. (2017), Verma
et al. (2016) and Eltawil and Zhao (2013)
• Online (no priori information is
required)
• Simple in logic and implementation
• Oscillations around MPP at steady state during fast change
in atmospheric conditions
• Suitable size for perturbation is important
• Less efficient in handling dynamic state
dc – dc
converter
PI Controller
D
Vref
VPV
VPV
Fig. 6. The block diagram of CV controller.
Table 6
Merits and demerits of CV technique.
Ref. Merits demerits
Tofoli et al. (2015), Gupta et al. (2016), Danandeh
and Mousavi (2017) and Eltawil and Zhao (2013)
• Simple, fast and easy to implement
• CV uses only one voltage sensor with no need for
the current sensor hence, the cost is less
• Economical and more efficient during low
radiation
• Offline (priori information is required)
• Less MPPT accuracy and efficiency due to approximation
(VMPP = 0.76%VO.C) that is not true in all cases
945

trained once for several PSCs to determine the global voltage (VGP). The
FLC uses VGP as a reference voltage to adjust the duty cycle of the boost
converter (Karatepe and Hiyama, 2009). The merits and demerits of
ANN based MPPT are listed in Table 11.
3.2.1.3. Adaptive neuro fuzzy inference system technique. Adaptive Neuro
Fuzzy Inference System (ANFIS) is one of the most efficient AI
techniques based MPPT that uses ANN for internal data training and
FLC for external data. Hence, it has the advantages of both techniques.
The inputs of ANN are error (E) and error (ΔE) and the ANN output will
be the input to FLC. The FLC provides the optimal duty cycle of the dc-
dc converter to track the GP (Saravanan and Babu, 2016). It is difficult
to obtain membership functions and fuzzy rules using trial and error, so,
the ANN part in ANFIS reduces the error and optimizes the parameters.
Whereas, FLC has the ability to work with imprecise inputs and good
efficiency in addition to accurate mathematical model and detailed
information of the system are not required (Gupta et al., 2016;
Belhachat and Larbes, 2017).
Based on recent comparative studies of ANFIS based MPPT shown in
Table 12, it can be observed that ANFIS can extract the maximum
power efficiently and accurately regardless of whether PSC occurs or
not. Radianto et al. proved that ANFIS can extract the GP of the TCT
configuration through adjusting the duty cycle of the boost converter
(Radianto et al., 2012). This is supported by Faiza et al. which revealed
that ANFIS can track the GP efficiently and accurately under various
configurations such as HC, BL, TCT and SP. In addition, the highest
maximum power has been achieved with TCT configuration (Belhachat
and Larbes, 2017). The merits and demerits of ANFIS based MPPT are
shown in Table 13.
3.2.1.4. Differential evolution and genetic algorithm. Differential
Evolution (DE) is one of the most powerful stochastic, optimizing
based evolutionary algorithms (EA) which is similar to GA. Unlike, GA
which relies on crossover, DE relies on mutation (difference vector) to
Table 8
Recent comparative studies of FLC with other MPPT techniques.
Ref. Year Variable control dc-dc converter Application Findings
Rezk and Eltamaly (2015) 2015 Duty cycle Boost Standalone FLC has better performance in terms of tracking speed and drift avoidance followed by P&
O, INC, and HC MPPT techniques in both dynamic response and steady state
Chen et al. (2016) 2016 Duty cycle Boost Standalone FLC based auto-scaling variable step-size is proposed to achieve the merits of fast tracking
and convergence speed during transient and steady state (No oscillations) compared to
fixed step IC in both simulation and experimental works
Boukenoui et al. (2016) 2016 Duty cycle Boost Standalone The proposed FLC with a scanning and storing algorithm has good performance compared
to variable step size IC, conventional PSO, and FLC based HC in both simulation and
experimental works. Many merits are achieved such as fast and accurate convergence to the
GP, high tracking efficiency, no oscillations during transient and steady state conditions
Nabipour et al. (2017) and
Kwan and Wu (2016)
2017
2016
Duty cycle Boost-SEPIC Standalone Adaptive FLC performed well compared to the direct and indirect FLC based MPPT in terms
of the active power and current oscillations, rising time, settling time and over/undershoots
during dynamic and steady state conditions. The antecedent and consequent membership
functions of the proposed adaptive FLC are tuned synchronously
Guenounou et al. (2014) 2014 Duty cycle Boost Standalone An adaptive gain FLC outperforms the conventional FLC where it integrates two different
rules. The first rule is used to adjust the duty cycle of the boost converter while the second
one is used for online adjusting of the controller’s gain
Bendib et al. (2014) 2014 Duty cycle Buck Standalone FLC has good performance compared to P&O during dynamic and steady state conditions in
terms of tracking efficiency and response time
Kermadi and Berkouk (2017) 2017 Duty cycle Buck-Boost Standalone The three best MPPT techniques are FLC, GA and PSO in terms of the performance (tracking
speed, the average tracking error, the variance and the efficiency) and the implementation
cost (sensors type, circuit type and software complexity). PID and ANN are lesser
performance
El Khateb et al. (2013) 2013 Duty cycle SEPIC Standalone FLC based MPPT performed well than P&O in terms of accuracy, tracking speed and
convergence speed during dynamic and steady state conditions in both simulation and
experimental works
Table 7
Fuzzy rules for the input and output variables.
ΔErr
Err NB NM NS ZE PS PM PB
NB NB NB NB NB NM NS ZE
NM NB NB NB NM NS ZE PS
NS NB NB NM NS ZE PS PM
ZE NB NM NS ZE PS PM PB
PS NM NS ZE PS PM PB PB
PM NS ZE PS PM PB PB PB
PB ZE PS PM PB PB PB PB
To
Switch
gate
Fuzzy Logic
Controller
Calculation of Err and ΔErr Subsystem
Err
ΔErr
Vpv
Ipv
V_PV
I_PV
ΔD D
Fig. 7. FLC block diagram in Matlab/Simulink.
-100 100
NB NM NS ZE PS PM PB
Input membership functions of Err and ΔErr respectively
Output membership functions
-50 50
Err, MFs
ΔErr, MFs
ΔD, MFs
Fig. 8. Membership functions of FLC.
946

convert the operating point towards the best solution in a search area
(Salam et al., 2013). It also depends on the generation of initial random
population similar to the other EA where it refines and improves the
further candidate solutions using selection, mutation, and crossover. On
the other hand, GA depends on the survival of the fittest through first,
generation of initial random population. Then, an objective function is
Table 10
Recent comparative studies of ANN with other MPPT techniques.
Messalti et al.
(2017)
2017 Duty cycle Flyback converter Standalone PV Simulation and experimental findings revealed that variable step size ANN has better
performance in terms of tracking accuracy, response time, overshoot and ripple compared to
the fixed step size ANN that has the same disadvantages of P&O technique related to tracking
speed and oscillations around MPP at steady state
Rizzo and Scelba
(2015)
2015 Duty cycle Boost Standalone PV ANN is used directly to track the GP while the P&O technique is used only to refine the
result. The prediction accuracy depends on the preselected number of power measurements,
the ANN size and prior information
Bouselham et al.
(2017)
2017 Duty cycle Boost Standalone PV ANN with a scanning and storing algorithm has good performance in terms of tracking
speed, response time and efficiency compared to variable P&O with global scanning and IC
based on FLC
Jiang et al. (2015) 2015 Duty cycle Buck-boost Standalone PV Two implementations of ANN are combined with P&O where ANN is used to predict the GP
searching area and P&O is used to track the GP. The proposed hybrid MPPT can track the GP
efficiently and accurately in terms of tracking speed and convergence speed compared to P&
O, Fibonacci search, conventional PSO and DE
Punitha et al.
(2013)
2013 VMPP Buck Standalone PV The proposed ANN combined with IC can track the GP efficiently compared to P&O and FLC
based HC in terms of tracking speed and convergence speed. An ANN is used to provide Vref
to the modified IC
Karatepe and
Hiyama (2009)
2009 Duty cycle Boost Standalone PV ANN is combined with FLC where the former is used to track the GP under several PSCs with
SP, BL and TCT configurations
Table 9
Merits and demerits of FLC based MPPT technique.
Kandemir et al. (2017), Seyedmahmoudian et al.
(2016), Ram et al. (2017), Danandeh and
Mousavi (2017), Gounden et al. (2009),
Bounechba et al. (2014) and Chi, (2010)
• Highly robust, fast response, better performance and adjustable
accuracy
• Less oscillation during conditions variation.
• Able to work with imprecise inputs and good efficiency
• More effective when combined with other EA techniques
• Does not require accurate mathematical model and detailed
information of the system.
• High complexity and expensive.
• Offline (priori information is required)
• Efficiency of the whole system is dependent on the
designer's performance and precision of the rules.
• Fails to converge under dynamic states.
• Rules cannot be changed, once defined.
Table 11
Merits and demerits of ANN based MPPT technique.
Kandemir et al. (2017), Seyedmahmoudian et al. (2016), Ram et al. (2017),
Danandeh and Mousavi (2017), Nabipour et al. (2017) and Anh (2014)
• Fast tracking speed, acceptable
accurateness
• Effective, less oscillations in conditions
variation and good efficiency
• Can be trained offline and used in the on-
line environment
• High complexity and expensive.
• Require extensive information about
the PV parameters
• Additional cost of temperature and
irradiance sensors.
Table 12
Recent comparative studies of ANFIS with other MPPT techniques.
Gupta et al. (2016) 2016 Duty cycle Boost Standalone PV Firstly, ANFIS has better tracking efficiency than FLC and ANN techniques. Secondly, IC has
superior performance compared to P&O and CV in terms of tracking efficiency, rise time,
fall time and dynamic response. Finally, Neural-FL has better efficiency than other
conventional and hybrid MPPT techniques
Aldair et al. (2017) 2017 Duty cycle Buck Standalone PV Design and implementation of the proposed ANFIS, CV and IC using Altera EP4CE6E22C8N
FPGA card. The findings revealed that ANFIS is more efficient and better dynamic response
followed by IC and finally CV
Belhachat and Larbes
(2017)
2016 Duty cycle Boost Standalone PV ANFIS can track the GP efficiently and accurately under various PSCs and different
configurations such as HC, BL, TCT and SP. TCT has the best performance with the highest
maximum power
Kharb et al. (2014) 2014 Duty cycle Boost Standalone PV ANFIS can track the unique MPP quickly and efficiently under dynamic and steady state
conditions
Radianto et al. (2012) 2012 Duty cycle Boost Standalone PV ANFIS is used to extract the GP of the TCT configuration through adjusting the duty cycle of
the boost converter
947

defined to determine the fitness of each solution. Followed by
evaluating the fitness of each individual and finally creating a new
population using genetic operators (selection, crossover, mutation)
(Kermadi and Berkouk, 2017).
Based on recent comparative studies of DE and GA with other MPPT
techniques introduced in Table 14, although DE has acceptable per-
formance to track GP with PS; but, some modifications and improve-
ments on DE have been done by Ramli et al. (2015) which showed that
DE performed more efficient in tracking the GP under PSC compared to
classic PSO in terms of accuracy, tracking speed, convergence speed and
efficiency where classic PSO may trap at LP for some PSCs.
The proposed MPPT technique shown in (Ramli et al., 2015) has
three main merits that are (1) No random numbers used, (2) One tuning
parameter required (mutation factor), and (3) Implementation simpli-
city (Ramli et al., 2015). In addition, a modified DE proposed by Ta-
juddin et al. outperformed the HC to track the GP in terms of con-
vergence speed, tracking speed and accuracy. In addition, oscillation
around MPP did not occur during dynamic and steady state conditions
(Tajuddin et al., 2013). Finally, Kumar et al. proposed that Jaya DE that
can track the GP accurately and quickly compared to the state of the art
improved P&O with ACO (ACOPO), PSO and FPA techniques in terms of
tracking speed, convergence speed and accuracy under dynamic and
steady state conditions (Kumar et al., 2017). On the other hand, GA also
has the ability to track the GP under PSC where the comparative study
achieved by Yousra et al. revealed that GA can track the GP under PSC
compared to conventional techniques (P&O and IC) which fail to detect
GP and track the first MPP whether it is GP or LP (Shaiek et al., 2013).
In addition, Ramaprabha et al. proved that both GA and the binary
search method can track the GP for all PS patterns efficiently and ac-
curately (Ramaprabha and Mathur, 2012). Also, a comparative study
done by Kermadi et al. revealed that, both GA and PSO track the GP
with good performance and less implementation cost whereas, design
and implementation of GA is more difficult and complex than PSO
(Kermadi and Berkouk, 2017). On the other hand, many researchers
proposed that, GA should be combined and optimized by other MPPT
technique due to GA may fall in LP in some cases of PSCs. For example,
GA is optimized and combined with P&O to improve the performance
and efficiency of GA in handling and catching the GP under PSC where
GA parameters are decreased and the GP was tracked in a shorter time
(Daraban et al., 2014). Also, GA combined with FLC can improve the
performance and efficiency of FLC where GA can obtain the best subsets
of the membership functions. Optimized FLC has better performance in
terms of tracking speed, response time and efficiency in addition to
robustness than FLC alone (Larbes et al., 2009). Finally, merits and
demerits of DE and GA based MPPT techniques are shown in Table 15.
3.2.2. Bio-Inspired based MPPT techniques
In recent years, numerous review papers (Ramli et al., 2017;
Ishaque and Salam, 2013; Liu et al., 2016; Kamarzaman and Tan, 2014;
Kandemir et al., 2017; Mohapatra et al., 2017; Salam et al., 2013;
Seyedmahmoudian et al., 2016) concentrated on the general descrip-
tions of conventional, AI, and Bio-Inspired (BI) MPPT techniques in-
dividually or collectively including idea of operation, literature review
and different classifications of MPPT techniques that can be easily
found. Therefore, this section will focus on the recent comparative
studies of the BI MPPT techniques to highlight and determine the most
effective and efficient BI based MPPT techniques during simulation or
experimental work or both.
Based on the recent comparative studies of BI MPPT techniques
introduced in Table 16, both Cuckoo Search Optimization (CSO) and
Table 13
Merits and demerits of ANFIS based MPPT techniques.
Gupta et al. (2016), Belhachat and Larbes
(2017)
• Higher efficiency under PSCs, faster tracking speed, and robustness
• Collect advantages of both FL and ANN.
• Simple and does not require too much computing or mathematical
equations.
• The ANN part in ANFIS reduces the error and optimizes the
parameters.
• High complexity and expensive
• It is difficult to obtain membership functions and rules
• More sensors are required.
• Insufficient training on the PV array leads to less
accuracy
Table 14
Recent comparative studies of DE and GA with other MPPT techniques.
Ramli et al. (2015) 2015 Duty cycle Buck-boost Standalone PV Modified DE outperformed classic PSO to track the GP under PSC in terms of accuracy,
tracking speed, convergence speed and efficiency. Classic PSO may trap at LP for some
PSCs
Tajuddin et al. (2013) 2013 Duty cycle Buck-boost Grid-Connected DE is proposed to study its effectiveness in handling PSCs (Variable GP). It outperformed
the HC in terms of convergence speed, tracking speed and accuracy. Also, no oscillation
around MPP occurred during dynamic and steady state
Kumar et al. (2017) 2017 Duty cycle Boost Standalone PV Jaya DE can track the GP accurately and quickly where it outperformed the state of the art
improved P&O with ACO (ACOPO), PSO and FPA techniques in terms of tracking speed,
convergence speed and accuracy under dynamic and steady state
Kermadi and Berkouk
(2017)
2017 Duty cycle Buck-Boost Standalone GA and PSO outperformed PID, FLC and ANN, in terms of the performance and the
implementation cost (sensors type, circuit type and software complexity). Both GA and
PSO provide a good GP tracking and show very good performance but, design and
hardware implementation of GA is more difficult and complex than PSO
Shaiek et al. (2013) 2013 Duty cycle Boost Standalone PV GA succeeded in tracking the GP under PS compared to P&O and IC which fail to detect GP
and track the first MPP whatever it is GP or LP
Ramaprabha and Mathur
(2012)
2012 Duty cycle Boost Standalone PV Both GA and the binary search method can track the GP in all PS cases efficiently and
accurately with error percentage less than 2%
Daraban et al. (2014) 2014 Duty cycle Buck Grid-Connected GA is integrated with P&O. The GA parameters (population size and number of iterations)
are decreased, thus catching the GP in a shorter time
Larbes et al. (2009) 2009 Duty cycle Boost Standalone PV GA combined FLC to improve the performance and efficiency of FLC where GA can
optimize the FLC membership functions and rules. It has better performance in terms of
tracking speed, response time, efficiency, and robustness
948

PSO can track the GP under PSC efficiently and accurately but CSO has
better performance compared to PSO in terms of convergence speed,
response time and tracking speed (Ahmed and Salam, 2014). In addi-
tion, modifications and improvements on PSO done by Prasanth et al.
(Leader PSO) (Ram and Rajasekar, 2017) or Ishaque et al. (Determi-
nistic PSO) (Ishaque et al., 2012) improve the PSO performance in
terms of tracking speed, convergence to GP and efficiency to be similar
in performance with Firefly Algorithm (FA) and CSO. On the other
hand, Teaching–learning-based optimization (TLBO) technique is pro-
posed and compared with the conventional PSO and FLC under PSCs.
The findings revealed that TLBO performed well be compared to PSO
and FLC in terms of tracking speed, convergence speed and average
tracking time of GP (Belhachat and Larbes, 2018).
As shown in Table 16, Sundareswaran et al. proved that ABC per-
formed better in tracking the GP under PSCs compared to PSO and
enhanced P&O (Sundareswaran et al., 2015). In addition, a modified
ABC (MABC) is proposed by Fathy (2015) for mitigating the PSC effect
and the findings revealed that MABC is the most efficient MPPT tech-
nique in handling and tracking the GP under PSC compared to GA, PSO
and ABC. On the other hand, Jiang et al. proposed a novel ACO to track
the GP under PSC. The findings proved that ACO converge faster, less
number of iterations required and performed well in tracking the GP
under various PSC compared to PSO (Jiang et al., 2013). Sundar-
eswaran et al. revealed that both FA and PSO converge to GP but the
convergence time for FA is shorter than that for PSO. Also, FA per-
formed well compared to PSO and P&O (Sundareswaran et al., 2014).
On the other hand, Teshome et al. proposed a modified FA (MFA) to
track the GP under PSCs and they claimed that it performed well
compared to FA in terms of tracking speed, convergence speed and
tracking efficiency. MFA can save 67% of the tracking time and track
the GP even under fast irradiance changes faster by 2 s compared with
FA (Teshome et al., 2017). Finally, Prasanth et al. proposed a new GP
tracking technique called Flower Pollination Algorithm (FPA) which
has lesser tracking time and faster convergence to the GP in all PSC
compared to PSO and P&O. The success behind FPA results from the
randomness in global and local search, two tuning parameters required
and very less computations compared to the required parameters of
PSO, ABC, FA, CSO and DPSO techniques (Ram and Rajasekar, 2017).
Based on the above discussion, some important conclusions can be
obtained. Firstly, most of BI based MPPT techniques have efficient
performance with and without PSC compared to conventional and AI
MPPT techniques. They can track the GP without falling or trapping in
LPs with high tracking speed, high convergence speed, less response
time and high efficiency. Secondly, modifications and improvements on
BI MPPT techniques improved their performance more and more in
terms of tracking speed, convergence speed and efficiency. Finally,
merits and demerits of all previous Bio-Inspired MPPT techniques are
introduced in Table 17.
Table 15
Merits and demerits of DE and GA based MPPT techniques.
Ref. Technique Merits Demerits
Seyedmahmoudian et al. (2016) and Ramli et al. (2015) DE • Simple and straightforward
• Rapid convergence
• Capable of tracking the GP regardless of the
initial parameter values
• Few control parameters required
• Slow convergence to the GP
• Limited local search ability
Seyedmahmoudian et al. (2016), Ram et al. (2017), Danandeh and Mousavi
(2017), Nabipour et al. (2017) and Fathy (2015)
GA • High speed, accuracy and good efficiency
• Possible wide search
• Applicable to fast change in atmospheric
conditions
• High complexity and
expensive.
• Much computation process
• High memory needed
• More time consumed
Table 16
Recent comparative studies of BI MPPT techniques.
Ahmed and Salam (2014) 2014 Duty cycle Boost Standalone PV Both CSO and PSO have high accuracy and stability in tracking GP but CSO has better
performance compared to PSO in terms of tracking speed and convergence speed
Ram and Rajasekar (2017) 2017 Duty cycle Boost Standalone PV Leader PSO has high tracking speed compared to PSO and P&O. It has similar performance
as DPSO and FA. Also, zero steady state oscillations and high convergence to GP is
achieved
Belhachat and Larbes
(2018)
2017 Duty cycle Boost Standalone PV TLBO is simple and has better performance compared to PSO and FLC in terms of tracking
speed, convergence speed and average tracking time of GP
Soufyane Benyoucef et al.
(2015)
2015 Duty cycle Boost Standalone PV ABC outperformed PSO in tracking the GP under PS and dynamic conditions. Also, it is
simple, uses fewer control parameters, convergence is independent of the initial conditions
and prior knowledge about the PV array is not required
Sundareswaran et al.
(2015)
2015 Duty cycle Boost Standalone PV ABC has faster tracking and less oscillation at steady state compared to PSO and Enhanced
P&O
Fathy (2015) 2015 Duty cycle Boost Standalone PV The findings revealed that the modified ABC is the most efficient in mitigating the power
loss under PS effect compared to GA, PSO and ABC
Jiang et al. (2013)) 2013 Duty cycle Boost Standalone PV ACO has faster convergence speed and requires less number of iterations to converge than
PSO. It is convergence independent of the initial conditions and provides better
performance to find the GP under various PSCs
Sundareswaran et al.
(2014)
2014 Duty cycle Boost Standalone PV FA is simple and has better performance compared to PSO and P&O in terms of tracking
speed, convergence speed and tracking efficiency. Both FA and PSO converge to GP but the
convergence time for FA is smaller than that for PSO
Teshome et al. (2017) 2017 Duty cycle two-phase IBC Standalone PV Modified FA (MFA) performed well compared to FA in terms of tracking speed,
convergence speed and tracking efficiency. MFA can save 67% of the tracking time and
track the GP even under fast irradiance changes faster than 2 s compared to FA
Ram and Rajasekar (2017) 2017 Duty cycle Boost Standalone PV FPA has fast convergence with lesser tracking time to catch the GP in all the PS cases
compared to PSO and P&O. FPA success is achieved due to the randomness in global and
local search and two tuning parameters required compared to the parameters required
PSO, ABC, FA, CSO and DPSO technique
949

3.3. Hybrid MPPT techniques
Hybrid MPPT techniques is a combination of conventional/soft
computing (Daraban et al., 2014) or soft computing/conventional
(Punitha et al., 2013; Jiang et al., 2015) or soft computing/soft com-
puting (Karatepe and Hiyama, 2009; Larbes et al., 2009; Kulaksız and
Akkaya, 2012) in order to handle the PSCs and track the GP accurately
and efficiently. Some soft computing MPPT especially those based on AI
such as FLC and ANN cannot handle the PSCs where they may fail to
track the GP. Therefore, they are optimized and integrated with other
techniques to improve the tracking efficiency and convergence speed.
Based on the recent combined and hybrid MPPT techniques in-
troduced in Table 18, soft computing based AI can be used to optimize
another soft computing based AI where GA optimized FLC membership
functions and rules as in (Larbes et al., 2009) while GA optimized ANN
as in (Kulaksız and Akkaya, 2012) and finally, ANN optimized FLC
performance where ANN trained under several PSCs to determine the
GP voltage and FLC uses the GP voltage to send the duty cycle to the
boost converter (Karatepe and Hiyama, 2009). On the other hand,
conventional P&O is embedded inside the soft computing GA to in-
crease the GA’s effectiveness for tracking the GP in a shorter time
through reducing the GA parameters (Daraban et al., 2014). The con-
ventional MPPT techniques are not accurate and stuck at the first peak
regardless of whether it is LP or GP. Therefore, soft computing can be
used to optimize the conventional one like ANN-IC where ANN provide
Vref to IC (Punitha et al., 2013) and ANN-P&O where ANN is used to
predict the GP region then P&O tracks the GP (Jiang et al., 2015).
4. Comparisons of all MPPT techniques
Based on the comparative studies presented previously for con-
ventional, AI and BI based MPPT techniques in addition to compre-
hensive review papers till the end of 2017, the authors only highlight
the best technique in terms of certain evaluation parameters such as
tracking speed and drift avoidance. However, there are many other
evaluation parameters which should also be taken into consideration
such as convergence speed, complexity, hardware implementation, in-
itial parameters required, performance with and without PS and effi-
ciency. As a result, technical and economical comparisons of conven-
tional, AI and BI based MPPT techniques based on 17 evaluation
parameters are shown in Tables 19, 20 and 21, respectively. Table 19
shows that the conventional MPPT techniques especially IC and P&O
followed by HC and CV are efficient, accurate and reliable to track the
unique MPP under uniform conditions (un-shaded) but they failed to
track the GP and stuck at the first MPP whatever it is GP or LP under
PSCs in terms of tracking speed, convergence speed, ability to track true
maxima and efficiency. Based on Table 20, the AI MPPT techniques
such as DE, ANFIS and GA are able to track the GP under PSCs with
medium tracking speed, convergence speed, ability to track true
maxima and efficiency while FLC and ANN have less ability to track the
GP. Therefore, they are optimized and combined with other techniques
to improve the tracking efficiency and convergence speed. On the other
hand, Table 21 shows that the BI MPPT techniques are more efficient,
accurate and reliable to track the GP instead of LPs under PSCs com-
pared to the other conventional and AI MPPT techniques.
5. Total evaluation of all MPPT techniques
Fig. 9 shows the comparison of 17 MPPT techniques under study
based on conventional, soft computing (AI and BI) according to the 8
most important evaluation parameters which are tracking speed, con-
vergence speed, complexity, hardware implementation, initial para-
meters required, performance without PS, performance with PS, and
efficiency from Not Applicable (NA), Very Low (V.L), Low (L), Medium
(M), High (H), and Very High (V.H). Each evaluation parameter has five
points and some evaluation parameters have positive trend such as
tracking speed, convergence speed, performance without PSC, perfor-
mance with PSC, and efficiency (NA = 0; V.L = 1; L = 2; M = 3; H = 4;
Table 17
Merits and demerits of bio-inspired MPPT techniques.
Ref. Technique Merits Demerits
Liu et al. (2016), Kandemir et al. (2017), Mohapatra et al.
(2017), Ram et al. (2017), Danandeh and Mousavi (2017),
Nabipour et al. (2017), Fathy (2015) and Miyatake et al.
(2011)
PSO • Efficient, accurate and fast tracking speed
• Simple, reliable and robustness
• Good performance with PS and does not depend on
PV arrays
• Highly effective in GP tracking
• No oscillations around MPP at steady state
• Initialization and computation are
difficult in large population
• Convergence cannot be achieved if
GP located outside the search area
Belhachat and Larbes (2018) TLBO • Simple and rapid tracking speed and convergence
• Tracking speed is improved than conventional PSO
Liu et al. (2016), Mohapatra et al. (2017), Ram et al. (2017) and
Ahmed and Salam (2014)
CSO • Efficient randomization, good robustness and rapid
tracking speed
• Does not require an accurate mathematical model
• Good transient performance and fast convergence
• higher efficiency using fewer tuning parameters
• No steady state oscillations around MPP
• Much complex computation
• Tracking time depends upon levy
flight
• Deterioration of convergence speed
and quality
Mohapatra et al. (2017), Seyedmahmoudian et al. (2016) and
Jiang et al. (2013)
ACO • Simple control, low cost, and robust
• Good performance with PS and does not depend on
PV arrays.
• Five initial parameters required
• Much and complex computation
Mohapatra et al. (2017), Danandeh and Mousavi (2017) and
Sundareswaran et al. (2014)
FA • Faster convergence and more accurate
• High efficiency, never fall on LPs
Mohapatra et al. (2017), Belhachat and Larbes (2017),
Sundareswaran et al. (2015), Fathy (2015) and Soufyane
Benyoucef et al. (2015)
ABC • Simple and fewer control parameters used
• Convergence is independent of initial conditions.
• PV Prior knowledge is not required
• Slow tracking and complex
• May fall on LPs because of fewer
control parameters
Ram and Rajasekar (2017) FPA • Robustness, fast convergence with lesser tracking time
to catch GP
• Simple in construction and implementation
• Updating the duty cycle done using two simple steps
(Cross pollination and Local pollination)
• High cost
Mohapatra et al. (2017) and Seyedmahmoudian et al. (2016) GWO • Higher tracking efficiency and fast convergence with
zero steady state oscillations
• Robust, reliable and fewer parameters required
• Large search space, high cost
950

V.H = 5 points). Whereas, the other ones have negative trend such as
complexity, hardware implementation and initial parameters required
(V.L = 5; L = 4; M = 3; H = 2; V.H = 1 points). For example, the total
evaluation of Flower Pollination Algorithm (FPA) is equal to 35/40
points (5 + 3 + 4 + 4 + 4 + 5 + 5 + 4) starting from tracking speed
and so on for all MPPT techniques. The idea behind this is to make total
and comprehensive evaluation to highlight and determine the most
efficient techniques with and without PSC.
Fig. 10 shows the total evaluation chart of conventional, AI and BI
MPPT techniques. This chart proves that, FPA represents the best MPPT
technique followed by FA based on total evaluation with 35 and 34
points from 40 points; respectively. This conclusion matches with the
finding that FPA performed well compared to PSO, ABC, FA and CSO
which was revealed by Prasanth in (Ram and Rajasekar, 2017). In ad-
dition, the most effective and efficient BI based MPPT techniques with
PSC are FPA, FA and CSO followed by GWO, ABC, PSO and ACO;
Table 18
Recent combined and hybrid MPPT techniques.
Boukenoui et al. (2016) 2016 Duty cycle Boost Standalone FLC combined with a scanning and storing algorithm has fast and accurate
convergence to the GP, high efficiency, no oscillations during transient and steady state
conditions compared to variable step size IC, conventional PSO, and FLC based HC in
both simulation and experimental works
Punitha et al. (2013) 2013 VMPP Buck Standalone PV Hybrid ANN-IC can track the GP under PSC effectively and accurately compared to P&
O and FLC based HC where ANN provide Vref to IC
Jiang et al. (2015) 2015 Duty cycle Buck-boost Standalone PV Hybrid ANN-P&O can track the GP efficiently and accurately compared to P&O,
Fibonacci search, conventional PSO and DE. ANN is used to predict the GP region then
P&O tracks the GP
Karatepe and Hiyama (2009) 2009 Duty cycle Boost Standalone PV ANN-FLC where ANN trained once under several PSCs to determine the GP voltage
with SP, BL and TCT configurations. FLC uses the GP voltage to send the duty cycle for
the boost converter
Daraban et al. (2014) 2014 Duty cycle Buck Grid-Connected Hybrid P&O-GA can catch the GP in a shorter time because the GA parameters
(population size and number of iterations) are decreased
Larbes et al. (2009) 2009 Duty cycle Boost Standalone PV Hybrid FLC-GA has better performance and efficiency compared to P&O where GA
optimized FLC membership functions and rules
Kulaksız and Akkaya (2012) 2012 – Not used Standalone PV GA optimized ANN based MPPT where GA is used to determine neuron numbers in
multi-layer perceptron neural network. The PV system design eliminate dc–dc
converter and its losses
Table 19
Comparison of conventional MPPT techniques.
References Jeddi and Ouni (2014), Tofoli et al.
(2015), Gupta et al. (2016), Verma et al.
(2016), Eltawil and Zhao (2013),
Danandeh and Mousavi (2017), Nabipour
et al. (2017) and Bendib et al. (2015)
Danandeh and Mousavi
(2017), Nabipour et al.
(2017), Verma et al. (2016)
and Eltawil and Zhao (2013)
Jeddi and Ouni (2014), Gupta et al.
(2016), Verma et al. (2016), Eltawil and
Zhao (2013), Subudhi and Pradhan
(2013), Danandeh and Mousavi (2017),
Nabipour et al. (2017) and Bendib et al.
(2015)
Tofoli et al. (2015), Gupta et al.
(2016), Danandeh and Mousavi
(2017), Verma et al. (2016) and
Eltawil and Zhao (2013)
Parameters P&O HC IC CV
Control strategy SM SM SM SM
Required sensors V, I V, I V, I V
Variable control Duty Cycle Duty Cycle Duty Cycle Duty Cycle
PV array dependency No No No Yes
Tracking speed Low Low Low Low
Convergence speed Low Low Low Low
Parameter tuning No No No No
Complexity Simple Simple Medium Simple
Hardware
implementation
Simple Simple Simple Simple
Analog/digital Both Both Digital Analog
Ability to track true
maxima
Poor Poor Poor Poor
Initial parameters
required
Yes Yes Yes Yes
Sensitivity Medium Low Medium Low
Performance without
PS
High High High Medium
Performance with PS NA NA NA NA
Efficiency Medium Low Medium Low
Cost Medium Medium Medium Low
951

respectively. In general, most BI MPPT techniques are efficient in
handling PSC and tracking the GP efficiently and accurately at all times
but a small difference occurs between them in some evaluation para-
meters such as convergence speed, complexity, hardware implementa-
tion and initialization parameters discriminates and makes the differ-
ence in their total evaluation as shown in Fig. 10. Both PSO and ACO
have the same total evaluation points but, optimization and improve-
ments that have been done on PSO may put it in high arrangement with
FPA and FA.
On the other hand, the most effective and efficient soft computing
MPPT techniques based AI with PSC are DE, ANFIS followed by GA but
they may not provide convergence to the GP in some cases of PSC (Ram
and Rajasekar, 2017). Although, both DE and ANFIS have the same
total evaluation points but, the improvements that have been done on
DE may put it in high ranking with MPPT techniques based BI. Both FLC
and ANN are not able to track the GP separately and fail to handle the
PSC since training ANN needs data which is not available from the
random nature of PSC. In addition, the membership function and con-
trol variables of FLC are static once defined, while PSC is dynamic and
tracking the varying GP is not a direct task (Ram et al., 2017). Also,
they require high computation controller ability for training in addition
to extensive information required about the PV system for training and
tracking rules (Ram and Rajasekar, 2017). Therefore, they should be
optimized and combined with other MPPT technique to improve their
performance and convergence to the GP with PSC. Finally, although
conventional MPPT techniques cannot be applied with PSC, they are
efficient in tracking the unique MPP under uniform conditions. Both IC
and P&O have good performance with the same total evaluation points
followed by HC and CV techniques. This conclusion matches with the
finding that some modifications and improvements on P&O put it in the
same performance level with IC.
6. Conclusions
This paper introduced a comparative and comprehensive review for
the 17 most famous MPPT techniques to track the GP instead of LPs and
alleviate the PSC effects. The 17 MPPT techniques are divided into three
groups which are conventional, soft computing (AI, BI) and hybrid
based MPPT techniques. One of the most significant findings introduced
in this study is the technical and economical assessment for the 17 most
famous and efficient MPPT techniques based on 17 evaluation para-
meters. In addition, a novel evaluation index has been introduced to
rank these 17 techniques based on the 8 most important key issues with
total evaluation by 40 points for these MPPT techniques. Ranking of
these 17 MPPT techniques will help researchers, scientists, and in-
dustrial sector to pick up the most effective and appropriate MPPT
technique easily. Based on these evaluations, both IC and P&O have the
same total evaluation points followed by HC and CV technique. This
conclusion matches with the previous literature finding revealed that
optimized P&O can have mostly the same efficiency as IC. Also, con-
ventional MPPT techniques are not applicable with PS where they can
track the unique peak efficiently and accurately but, they fail to track
the GP and stuck at the first MPP whatever it is GP or LP. Whereas,
MPPT techniques based AI are efficient and accurate to track the unique
peak under uniform condition compared to conventional techniques
and mitigate their shortcomings related to tracking speed and oscilla-
tions around MPP at steady state. The most effective and efficient AI
based MPPT techniques with PSC are DE, ANFIS followed by GA.
Although, both DE and ANFIS have the same total evaluation points
but, optimization and improvements that have been done in recent
years on DE may put it in high ranking with MPPT techniques based BI.
Both FLC and ANN are not able to track the GP separately and fail to
handle the PSC. Therefore, it should be optimized and combined with
Table 20
Comparison of AI based MPPT techniques.
Ref. Seyedmahmoudian et al.
(2016), Kichou et al.
(2016), Ishaque and
Salam (2011) and
Peñuñuri et al. (2016)
Seyedmahmoudian et al. (2016),
Danandeh and Mousavi (2017),
Nabipour et al. (2017), Kichou
et al. (2016), Bakhshi et al.
(2014) and Zhang et al. (2015)
Seyedmahmoudian et al.
(2016), Gupta et al. (2016),
Danandeh and Mousavi (2017),
Bendib et al. (2015), Verma
et al. (2016), Eltawil and Zhao
(2013), Gounden et al. (2009),
Bounechba et al. (2014) and
Chiu (2010)
Seyedmahmoudian et al. (2016),
Gupta et al. (2016), Verma et al.
(2016), Eltawil and Zhao (2013),
Anh (2014), Danandeh and
Mousavi (2017), Nabipour et al.
(2017) and Bendib et al. (2015)
Gupta et al. (2016),
Nabipour et al. (2017)
and Belhachat and
Larbes (2017)
Parameters DE GA FLC ANN ANFIS
Control strategy AI AI AI AI AI
Required sensors V, I V, I Depends Depends Depends
Variable control Duty Cycle Duty Cycle Duty Cycle Duty Cycle Output power
PV array dependency No No Yes Yes Yes
Tracking speed Medium Medium Medium Medium Medium
Convergence speed Medium Medium Medium Medium Medium
Parameter tuning No No Yes Yes No
Complexity High Medium Medium High High
Hardware
implementation
Medium High Medium High Medium
Analog/digital Digital Digital Digital Digital Digital
maxima
Medium High Low Poor Poor Medium High
Initial parameters
required
3 4 1 2 1
Sensitivity High Medium Medium Medium Medium
Performance without
PS
High High High High High
Performance with PS Medium Medium Low Low Medium
Efficiency Medium High Medium Medium Medium Medium High
Cost High High High High High
952

Table 21
Comparison of bio-inspired MPPT techniques.
Ref. Seyedmahmoudian et al. (2016),
Danandeh and Mousavi (2017), Nabipour
et al. (2017), Verma et al. (2016),
Sundareswaran et al. (2015), Fathy (2015)
and Kichou et al. (2016)
Belhachat and
Larbes (2018)
Liu et al.
(2016) and
Shi et al.
(2016)
(2016), Verma et al. (2016)
and Jiang et al. (2013)
(2016), Danandeh and Mousavi
(2017) and Sundareswaran
et al. (2014)
Sundareswaran et al. (2015),
Fathy (2015), Soufyane
Benyoucef et al. (2015) and
Kichou et al. (2016)
Ram and
Rajasekar
(2017)
Seyedmahmoudian
et al.(2016)
Evaluation
parameters
PSO TLBO CSO ACO FA ABC FPA GWO
Control strategy BI BI BI BI BI BI BI BI
Required sensors V, I V, I V, I V, I V, I V, I V, I V, I
Variable control Duty Cycle Duty Cycle Duty Cycle Duty Cycle Duty Cycle Duty Cycle Duty Cycle Duty Cycle
PV array dependency NO NO NO NO NO NO NO NO
Tracking speed Fast Fast V. Fast Fast Fast Fast V. Fast V. Fast
Convergence speed Medium Medium Fast Fast Fast Fast Fast Medium
Parameter tuning No No No No No No No No
Complexity Medium Simple Simple High Simple High Simple Medium
Hardware
implementation
Medium Medium Medium Medium Easy Medium Easy Medium
Analog/digital Digital Digital Digital Digital Digital Digital Digital Digital
maxima
Medium High High High High High High High High
Initial parameters
required
5 3 4 5 2 4 2 4
Sensitivity High High High High High High High High
Performance without
PS
V. High V. High V. High V. High V. High V. High V. High V. High
Performance with PS High High V. High High V. High V. High V. High V. High
Efficiency High Medium High High High High High High High
Cost High High High High High High High High
A.M.
Eltamaly
et
al.
Solar
Energy
174
(2018)
940–956
953

other techniques to improve their tracking efficiency. Finally, the most
obvious findings are that the MPPT techniques based BI are more effi-
cient, accurate and reliable to track the GP under PSCs compared to the
other conventional and AI based MPPT techniques. The most effective
and efficient three BI based MPPT techniques with PSC are FPA, FA and
CSO followed by GWO, ABC, PSO and ACO that provide convergence to
the GP at all times. Although, both PSO and ACO have the same total
evaluation points, the improvements that have been done on PSO may
put it in high ranking level with FPA and FA.
Conflict of Interest
The authors declared that there is no conflict of interest.
Acknowledgment
The authors extend their appreciation to the Deanship of Scientific
Research at King Saud University, Riyadh, Saudi Arabia for funding this
work through research group No (RG-1439-66).
References
Abdelsalam, A.K., Massoud, A.M., Ahmed, S., Enjeti, P.N., 2011. High-performance
adaptive perturb and observe MPPT technique for photovoltaic-based microgrids.
IEEE Trans. Power Electron. 26 (4), 1010–1021.
Ahmed, J., Salam, Z., 2014. A Maximum Power Point Tracking (MPPT) for PV system
using Cuckoo Search with partial shading capability. Appl. Energy 119, 118–130.
Ahmed, J., Salam, Z., 2015. An improved perturb and observe (P&O) maximum power
point tracking (MPPT) algorithm for higher efficiency. Appl. Energy 150, 97–108.
Aldair, A.A., Obed, A.A., Halihal, A.F., 2017. Design and implementation of ANFIS-re-
ference model controller based MPPT using FPGA for photovoltaic system. Renew.
Sustain. Energy Rev.
Anh, H.P.H., 2014. Implementation of supervisory controller for solar PV microgrid
system using adaptive neural model. Int. J. Electr. Power Energy Syst. 63,
1023–1029.
Ansari, M.F., Chatterji, S., Iqbal, A., 2010. A fuzzy logic control scheme for a solar pho-
tovoltaic system for a maximum power point tracker. Int. J. Sustain. Energ. 29 (4),
245–255.
Arunkumari, T., Indragandhi, V., 2017. An overview of high voltage conversion ratio DC-
DC converter configurations used in DC micro-grid architectures. Renew. Sustain.
Energy Rev. 77, 670–687.
Azzouzi, M., 2012. Comparaison between MPPT P&O and MPPT fuzzy controls in opti-
mizing the photovoltaic generator. Int. J. Adv. Comput. Sci. Appl. 3 (12), 57–62.
Babu, T.S., Rajasekar, N., Sangeetha, K., 2015. Modified particle swarm optimization
technique based maximum power point tracking for uniform and under partial
shading condition. Appl. Soft Comput. 34, 613–624.
Bakhshi, R., Sadeh, J., Mosaddegh, H.-R., 2014. Optimal economic designing of grid-
connected photovoltaic systems with multiple inverters using linear and nonlinear
module models based on Genetic Algorithm. Renew. Energy 72, 386–394.
Belhachat, F., Larbes, C., 2017. Global maximum power point tracking based on ANFIS
approach for PV array configurations under partial shading conditions. Renew.
Sustain. Energy Rev. 77, 875–889.
Belhachat, F., Larbes, C., 2018. A review of global maximum power point tracking
techniques of photovoltaic system under partial shading conditions. Renew. Sustain.
Energy Rev. 92, 513–553 2018/09/01/2018.
Bendib, B., Krim, F., Belmili, H., Almi, M., Boulouma, S., 2014. Advanced Fuzzy MPPT
Controller for a stand-alone PV system. Energy Procedia 50, 383–392.
Bendib, B., Belmili, H., Krim, F., 2015. A survey of the most used MPPT methods:
Conventional and advanced algorithms applied for photovoltaic systems. Renew.
Boukenoui, R., Salhi, H., Bradai, R., Mellit, A., 2016. A new intelligent MPPT method for
stand-alone photovoltaic systems operating under fast transient variations of shading
patterns. Sol. Energy 124, 124–142.
Bounechba, H., Bouzid, A., Nabti, K., Benalla, H., 2014. Comparison of perturb & observe
and fuzzy logic in maximum power point tracker for PV systems. Energy Procedia 50,
677–684.
Bouselham, L., Hajji, M., Hajji, B., Bouali, H., 2017. A new MPPT-based ANN for pho-
tovoltaic system under partial shading conditions. Energy Procedia 111, 924–933.
Cavalcanti, M., Oliveira, K., Azevedo, G., Neves, F., 2007. Comparative study of max-
imum power point tracking techniques for photovoltaic systems. Eletrônica de
Potência 12 (2), 163–171.
Chekired, F., Mellit, A., Kalogirou, S., Larbes, C., 2014. Intelligent maximum power point
trackers for photovoltaic applications using FPGA chip: A comparative study. Sol.
Energy 101, 83–99.
Chen, Y.-T., Jhang, Y.-C., Liang, R.-H., 2016. A fuzzy-logic based auto-scaling variable
step-size MPPT method for PV systems. Sol. Energy 126, 53–63.
Chiu, C.-S., 2010. TS fuzzy maximum power point tracking control of solar power gen-
eration systems. IEEE Trans. Energy Convers. 25 (4), 1123–1132.
Danandeh, M., Mousavi, S.G., 2017. Comparative and comprehensive review of maximum
power point tracking methods for PV cells. Renew. Sustain. Energy Rev.
Daraban, S., Petreus, D., Morel, C., 2014. A novel MPPT (maximum power point tracking)
algorithm based on a modified genetic algorithm specialized on tracking the global
maximum power point in photovoltaic systems affected by partial shading. Energy
74, 374–388.
El Khateb, A.H., Rahim, N.A., Selvaraj, J., 2013. Fuzzy logic control approach of a
maximum power point employing SEPIC converter for standalone photovoltaic
system. Procedia Environ. Sci. 17, 529–536.
Elgendy, M.A., Zahawi, B., Atkinson, D.J., 2012. Assessment of perturb and observe MPPT
algorithm implementation techniques for PV pumping applications. IEEE Trans.
Sustain. Energy 3 (1), 21–33.
Eltawil, M.A., Zhao, Z., 2013. MPPT techniques for photovoltaic applications. Renew.
Faranda, R., Leva, S., 2008. A Comparative Study of MPPT techniques for PV Systems. In:
7th WSEAS International Conference on Application of Electrical Engineering
(AEE’08), Trondheim, Norway.
Fathy, A., 2015. Reliable and efficient approach for mitigating the shading effect on
photovoltaic module based on Modified Artificial Bee Colony algorithm. Renew.
Energy 81, 78–88.
Femia, N., Petrone, G., Spagnuolo, G., Vitelli, M., 2005. Optimization of perturb and
Convergence
speed Complexity
Initial parameters
required
Performence
without PS
Performence
with PS
Efficiency
Tracking
speed
Hardware
implementation
V. L
L
M
H
V. H
FPA - CSO - GWO
PSO - TLBO - ACO-
FA - ABC
DE - GA - FLC -
ANN - ANFIS
P&O - HC - INC - CV
PSO - TLBO - GWO
CSO - ACO - FA
ABC - FPA
ACO-ABC-DE-ANN-ANFIS
TLBO - CSO - FA - FPA
P&O - HC - CV
PSO - GWO - GA - FLC - IC
FA - FPA
P&O - HC - IC - CV
PSO - TLBO - CSO
ACO - ABC - GWO
DE - FLC - ANFIS
2
3
4
5
2
3
4
5 PSO - ACO
CSO - ABC - GWO - GA
TLBO - DE
FA - FPA - ANN
Bio-Inspired Techniques
NA
Conventional Techniques
CSO - FA - ABC -FPA - GWO
PSO - CSO - ACO - FA
ABC - FPA - GWO
TLBO - DE - ANFIS
HC - CV
PSO - TLBO - ACO
Artificial Intelligent &
Conventional Techniques
P&O - HC - IC - CV
GA-ANN
GA - FLC - ANN
P&O - IC
DE - ANFIS - GA
FLC - ANN
1
1 FLC - ANFIS
Fig. 9. Comparisons of AI, BI and conventional MPPT techniques.
27 29.5
32
27
34
29
35
30
26.5
25
24
23
26.5
24
23
24 21
0
10
20
30
40
PSO
TLBO
CSO
ACO
FA
ABC
FPA
GWO
DE
GA
FLC
ANN
ANFIS
P&O
HC
IC
CV
Total Evaluation = 40 Points
Fig. 10. Total evaluation of the 17 MPPT techniques.
954

observe maximum power point tracking method. IEEE Trans. Power Electron. 20 (4),
963–973.
Ghassami, A.A., Sadeghzadeh, S.M., Soleimani, A., 2013. A high performance maximum
power point tracker for PV systems. Int. J. Electr. Power Energy Syst. 53, 237–243.
Gounden, N.A., Peter, S.A., Nallandula, H., Krithiga, S., 2009. Fuzzy logic controller with
MPPT using line-commutated inverter for three-phase grid-connected photovoltaic
systems. Renew. Energy 34 (3), 909–915.
Guenounou, O., Dahhou, B., Chabour, F., 2014. Adaptive fuzzy controller based MPPT for
photovoltaic systems. Energy Convers. Manage. 78, 843–850.
Gupta, A., Chauhan, Y.K., Pachauri, R.K., 2016. A comparative investigation of maximum
power point tracking methods for solar PV system. Sol. Energy 136, 236–253.
Hohm, D., Ropp, M.E., 2003. Comparative study of maximum power point tracking al-
gorithms. Prog. Photovoltaics Res. Appl. 11 (1), 47–62.
Houssamo, I., Locment, F., Sechilariu, M., 2010. Maximum power tracking for photo-
voltaic power system: Development and experimental comparison of two algorithms.
Renew. Energy 35 (10), 2381–2387.
Ishaque, K., Salam, Z., 2011. An improved modeling method to determine the model
parameters of photovoltaic (PV) modules using differential evolution (DE). Sol.
Energy 85 (9), 2349–2359.
Ishaque, K., Salam, Z., Amjad, M., Mekhilef, S., 2012. An improved particle swarm op-
timization (PSO)–based MPPT for PV with reduced steady-state oscillation. IEEE
Trans. Power Electron. 27 (8), 3627–3638.
Ishaque, K., Salam, Z., 2013. A review of maximum power point tracking techniques of PV
system for uniform insolation and partial shading condition. Renew. Sustain. Energy
Rev. 19, 475–488.
Ishaque, K., Salam, Z., Lauss, G., 2014. The performance of perturb and observe and
incremental conductance maximum power point tracking method under dynamic
weather conditions. Appl. Energy 119, 228–236.
Jeddi, N., Ouni, L.E.A., 2014. Comparative study of MPPT techniques for PV control
systems. In: Electrical Sciences and Technologies in Maghreb (CISTEM), 2014
International Conference on. IEEE, pp. 1–7.
Jiang, L.L., Maskell, D.L., Patra, J.C., 2013. A novel ant colony optimization-based
maximum power point tracking for photovoltaic systems under partially shaded
conditions. Energy Build. 58, 227–236.
Jiang, L.L., Nayanasiri, D., Maskell, D.L., Vilathgamuwa, D., 2015. A hybrid maximum
power point tracking for partially shaded photovoltaic systems in the tropics. Renew.
Energy 76, 53–65.
Kamarzaman, N.A., Tan, C.W., 2014. A comprehensive review of maximum power point
tracking algorithms for photovoltaic systems. Renew. Sustain. Energy Rev. 37,
585–598.
Kandemir, E., Cetin, N.S., Borekci, S., 2017. A comprehensive overview of maximum
power extraction methods for PV systems. Renew. Sustain. Energy Rev. 78, 93–112.
Karami, N., Moubayed, N., Outbib, R., 2017. General review and classification of different
MPPT Techniques. Renew. Sustain. Energy Rev. 68, 1–18.
Karatepe, E., Hiyama, T., 2009. Artificial neural network-polar coordinated fuzzy con-
troller based maximum power point tracking control under partially shaded condi-
tions. IET Renew. Power Gener. 3 (2), 239–253.
Kermadi, M., Berkouk, E.M., 2017. Artificial intelligence-based maximum power point
tracking controllers for Photovoltaic systems: Comparative study. Renew. Sustain.
Energy Rev. 69, 369–386.
Khadidja, S., Mountassar, M., M’hamed, B., 2017. Comparative study of incremental
conductance and perturb & observe MPPT methods for photovoltaic system. In: Green
Energy Conversion Systems (GECS), 2017 International Conference on. IEEE, pp. 1–6.
Kharb, R.K., Shimi, S., Chatterji, S., Ansari, M.F., 2014. Modeling of solar PV module and
maximum power point tracking using ANFIS. Renew. Sustain. Energy Rev. 33,
602–612.
Kichou, S., Silvestre, S., Guglielminotti, L., Mora-López, L., Muñoz-Cerón, E., 2016.
Comparison of two PV array models for the simulation of PV systems using five
different algorithms for the parameters identification. Renew. Energy 99, 270–279.
Koutroulis, E., Kalaitzakis, K., Voulgaris, N.C., 2001. Development of a microcontroller-
based, photovoltaic maximum power point tracking control system. IEEE Trans.
Power Electron. 16 (1), 46–54.
Kulaksız, A.A., Akkaya, R., 2012. A genetic algorithm optimized ANN-based MPPT al-
gorithm for a stand-alone PV system with induction motor drive. Sol. Energy 86 (9),
2366–2375.
Kumar, N., Hussain, I., Singh, B., Panigrahi, B., 2017. Rapid MPPT for uniformly and
partial shaded PV System by using JayaDE algorithm in highly fluctuating atmo-
spheric conditions. IEEE Trans. Ind. Inf.
Kwan, T.H., Wu, X., 2016. Maximum power point tracking using a variable antecedent
fuzzy logic controller. Sol. Energy 137, 189–200.
Larbes, C., Cheikh, S.A., Obeidi, T., Zerguerras, A., 2009. Genetic algorithms optimized
fuzzy logic control for the maximum power point tracking in photovoltaic system.
Renew. Energy 34 (10), 2093–2100.
Laudani, A., Fulginei, F.R., Salvini, A., Lozito, G., Mancilla-David, F., 2014.
Implementation of a neural MPPT algorithm on a low-cost 8-bit microcontroller. In:
Power Electronics, Electrical Drives, Automation and Motion (SPEEDAM), 2014
International Symposium on. IEEE, pp. 977–981.
Liu, L., Meng, X., Liu, C., 2016. A review of maximum power point tracking methods of
PV power system at uniform and partial shading. Renew. Sustain. Energy Rev. 53,
1500–1507.
Mahamudul, H., Saad, M., Ibrahim Henk, M., 2013. Photovoltaic system modeling with
fuzzy logic based maximum power point tracking algorithm. Int. J. Photoenergy
2013.
Mancilla-David, F., Riganti-Fulginei, F., Laudani, A., Salvini, A., 2014. A neural network-
based low-cost solar irradiance sensor. IEEE Trans. Instrum. Meas. 63 (3), 583–591.
Messai, A., Mellit, A., Guessoum, A., Kalogirou, S., 2011. Maximum power point tracking
using a GA optimized fuzzy logic controller and its FPGA implementation. Sol. Energy
85 (2), 265–277.
Messalti, S., Harrag, A., Loukriz, A., 2017. A new variable step size neural networks MPPT
controller: Review, simulation and hardware implementation. Renew. Sustain.
Energy Rev. 68, 221–233.
Miyatake, M., Veerachary, M., Toriumi, F., Fujii, N., Ko, H., 2011. Maximum power point
tracking of multiple photovoltaic arrays: A PSO approach. IEEE Trans. Aerosp.
Electron. Syst. 47 (1), 367–380.
Mohapatra, A., Nayak, B., Das, P., Mohanty, K.B., 2017. A review on MPPT techniques of
PV system under partial shading condition. Renew. Sustain. Energy Rev. 80,
854–867.
Nabipour, M., Razaz, M., Seifossadat, S.G., Mortazavi, S., 2017. A new MPPT scheme
based on a novel fuzzy approach. Renew. Sustain. Energy Rev. 74, 1147–1169.
Noman, A.M., Addoweesh, K.E., Mashaly, H.M., 2012. A fuzzy logic control method for
MPPT of PV systems. In: IECON 2012-38th Annual Conference on IEEE Industrial
Electronics Society. IEEE, pp. 874–880.
Pandey, A., Dasgupta, N., Mukerjee, A.K., 2008. High-performance algorithms for drift
avoidance and fast tracking in solar MPPT system. IEEE Trans. Energy Convers. 23
(2), 681–689.
Peñuñuri, F., Cab, C., Carvente, O., Zambrano-Arjona, M.A., Tapia, J., 2016. A study of
the Classical Differential Evolution control parameters. Swarm Evol. Comput. 26,
86–96.
Punitha, K., Devaraj, D., Sakthivel, S., 2013. Artificial neural network based modified
incremental conductance algorithm for maximum power point tracking in photo-
voltaic system under partial shading conditions. Energy 62, 330–340.
Radianto, D., Asfani, D.A., Hiyama, T., 2012. Partial shading detection and mppt con-
troller for total cross tied photovoltaic using anfis. ACEEE Int. J. Electr. Power Eng.
3 (2).
Rai, A.K., Kaushika, N., Singh, B., Agarwal, N., 2011. Simulation model of ANN based
maximum power point tracking controller for solar PV system. Sol. Energy Mater. Sol.
Cells 95 (2), 773–778.
Ram, J.P., Rajasekar, N., 2017. A new global maximum power point tracking technique
for solar photovoltaic (PV) system under partial shading conditions (PSC). Energy
118, 512–525.
Ram, J.P., Rajasekar, N., 2017. A new robust, mutated and fast tracking LPSO method for
solar PV maximum power point tracking under partial shaded conditions. Appl.
Energy 201, 45–59.
Ram, J.P., Babu, T.S., Rajasekar, N., 2017. A comprehensive review on solar PV maximum
power point tracking techniques. Renew. Sustain. Energy Rev. 67, 826–847.
Ramaprabha, R., Mathur, B., 2012. Genetic algorithm based maximum power point
tracking for partially shaded solar photovoltaic array. Int. J. Res. Rev. Informat. Sci.
(IJRRIS) 2.
Ramli, M.A., Ishaque, K., Jawaid, F., Al-Turki, Y.A., Salam, Z., 2015. A modified differ-
ential evolution based maximum power point tracker for photovoltaic system under
partial shading condition. Energy Build. 103, 175–184.
Ramli, M.A., Twaha, S., Ishaque, K., Al-Turki, Y.A., 2017. A review on maximum power
point tracking for photovoltaic systems with and without shading conditions. Renew.
Rezk, H., Eltamaly, A.M., 2015. A comprehensive comparison of different MPPT techni-
ques for photovoltaic systems. Sol. Energy 112, 1–11.
Rizzo, S.A., Scelba, G., 2015. ANN based MPPT method for rapidly variable shading
conditions. Appl. Energy 145, 124–132.
Salam, Z., Ahmed, J., Merugu, B.S., 2013. The application of soft computing methods for
MPPT of PV system: A technological and status review. Appl. Energy 107, 135–148.
Salas, V., Olias, E., Barrado, A., Lazaro, A., 2006. Review of the maximum power point
tracking algorithms for stand-alone photovoltaic systems. Sol. Energy Mater. Sol.
Cells 90 (11), 1555–1578.
Saravanan, S., Babu, N.R., 2016. Maximum power point tracking algorithms for photo-
voltaic system–A review. Renew. Sustain. Energy Rev. 57, 192–204.
Seyedmahmoudian, M., et al., 2016. State of the art artificial intelligence-based MPPT
techniques for mitigating partial shading effects on PV systems–A review. Renew.
Shaiek, Y., Smida, M.B., Sakly, A., Mimouni, M.F., 2013. Comparison between conven-
tional methods and GA approach for maximum power point tracking of shaded solar
PV generators. Sol. Energy 90, 107–122.
Sharma, R.S., Katti, P., 2017. Perturb & observation MPPT algorithm for solar photo-
voltaic system. In: Circuit, Power and Computing Technologies (ICCPCT), 2017
International Conference on. IEEE, pp. 1–6.
Shi, J.-Y., Xue, F., Qin, Z.-J., Zhang, W., Ling, L.-T., Yang, T., 2016. Improved global
maximum power point tracking for photovoltaic system via cuckoo search under
partial shaded conditions. J. Power Electron. 16 (1), 287–296.
Shimizu, T., Hashimoto, O., Kimura, G., 2003. A novel high-performance utility-inter-
active photovoltaic inverter system. IEEE Trans. Power Electron. 18 (2), 704–711.
Soufyane Benyoucef, A., Chouder, A., Kara, K., Silvestre, S., 2015. Artificial bee colony
based algorithm for maximum power point tracking (MPPT) for PV systems operating
under partial shaded conditions. Appl. Soft Comput. 32, 38–48.
Subudhi, B., Pradhan, R., 2013. A comparative study on maximum power point tracking
techniques for photovoltaic power systems. IEEE Trans. Sustainable Energy 4 (1),
89–98.
Sundareswaran, K., Peddapati, S., Palani, S., 2014. MPPT of PV systems under partial
shaded conditions through a colony of flashing fireflies. IEEE Trans. Energy Convers.
29 (2), 463–472.
Sundareswaran, K., Sankar, P., Nayak, P., Simon, S.P., Palani, S., 2015. Enhanced energy
output from a PV system under partial shaded conditions through artificial bee
colony. IEEE Trans. Sustain. Energy 6 (1), 198–209.
Tajuddin, M.F.N., Ayob, S.M., Salam, Z., Saad, M.S., 2013. Evolutionary based maximum
955

A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques

A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques

Recommended

Recommended

More Related Content

Similar to A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques

Similar to A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques (20)

More from Kathryn Patel

More from Kathryn Patel (20)

Recently uploaded

Recently uploaded (20)

A Novel Evaluation Index For The Photovoltaic Maximum Power Point Tracker Techniques