Professional Documents
Culture Documents
net/publication/332226932
CITATIONS READS
124 8,804
3 authors, including:
Some of the authors of this publication are also working on these related projects:
Hybrid system using solar cell, fuel cell and backup battery View project
All content following this page was uploaded by Amit Kumer Podder on 21 June 2019.
Review Article
Abstract: An efficient maximum power point tracking (MPPT) method plays an important role to improve the efficiency of a
photovoltaic (PV) generation system. This study provides an extensive review of the current status of MPPT methods for PV
systems which are classified into eight categories. The categorisation is based on the tracking characteristics of the discussed
methods. The novelty of this study is that it focuses on the key characteristics and eleven selection parameters of the methods
to make a comprehensive analysis, which is not considered together in any review works so far. Again, the pros and cons,
classification and immense comparison among them described in this study can be used as a reference to address the gaps for
further research in this field. A comparative review in tabular form is also presented at the end of the discussion of each
category to evaluate the performance of these methods, which will help in selecting the appropriate technique for any specific
application.
1 Introduction order term in the discrete differentiation of the power. The sliding
control (SC) method, however, complex in equipment usage yet, is
The steady increase in the level of greenhouse gas emissions and more precise than ordinary methods [14]. The classical algorithms,
fuel costs are the main motive behind the attempt to use different such as P&O, INC, HC, fuzzy logic and neural network, cannot
sources of renewable energy [1, 2]. Among various sustainable find the global MPP (GMPP) under partial shading condition
sources of energy, the solar energy is a suitable one because it is (PSC) [15]. A comparison among various global MPPT (GMPPT)
clean, free from emission and easy to change directly to electricity methods based on meta-heuristic algorithms is given in [16]. It is
utilising a photovoltaic (PV) system [2–4]. The generation of PV concluded that particle swarm optimisation (PSO) and Cuckoo
power has demonstrated a noteworthy potential in satisfying the search (CuS) algorithms based trackers ensure the convergence to
demand for energy. Up to the year 2016, the worldwide operation the GMPP and the tracking performance of the CuS algorithm is
of the sun-oriented power generation capacity has ascended to 302 better than the PSO. For effectively tracking the MPPT of a PV
GWp, which is enough to supply 1.8 per cent of the world energy system, a model-free spline-guided Jaya algorithm is proposed in
demand. The solar power generation capacity has increased by [17], which is able to perform efficiently under PSC and also
nearly 100 GWp in 2017, which is about 31 per cent more from provides faster convergence speed. An MPPT technique based on
2017 [5, 6]. However, the extensive use of a PV system is not so temperature described in [18] needs a fewer number of sensors
common because of its high starting cost. Again, there is no than customary strategies. This technique is straightforward in
assurance that the energy delivered from PV exhibits steady output execution and is economical too. The bisection search theorem-
since it relies completely on the sun-oriented irradiance and the based MPPT, detailed in [19], is generally utilised when the PV
surrounding temperature of the PV modules, cell region, and load. array shows at least two neighbourhood MPPs under changing
For efficient operation of the PV cell under prevailing climatic climatic conditions, where the utilisation of different methods is a
conditions, an appropriate mechanism is necessary for achieving troublesome undertaking.
maximum power from it, which is considered as a maximum power Until now, the operation of various MPPT methods is presented
point tracking (MPPT) in the literature. The MPPT increases the by various research papers. A lot of research works have also been
efficiency and lifetime of the PV module [7, 8]. published to classify these methods. Research works in [2] present
Researchers around the world create various methodologies to a categorisation approach for MPPT methods based on three
take out as much power as could reasonably be expected from categories such as off-line, online and hybrid methods. The
sustainable power sources and particularly, from the PV panels. research work in [20] categorises the MPPT methods as analogue,
Until now, a large number of MPPT algorithms are accessible in digital, and hybrid methods and provides a comparison based on
the literature for both off-grid and grid associated PV systems [9]. only five selection parameters. In [21], the MPPT methods are
The selection of a specific MPPT system from the various existing categorised into five categories based on their tracking techniques
MPPT methods is a confounding errand since every method has and the comparison that is given based on five selection
certain focal points and disadvantages [10]. For example, the hill parameters. These available categorisation methods cannot classify
climbing (HC) [11] and perturb and observe (P&O) [1, 2] methods all the available methods appropriately and the available
are broadly utilised as MPPT algorithms because of their simple comparison among the MPPT methods does not consider all the
execution and fewer sensor necessities. The incremental available selection parameters. An endeavour is made in this study
conductance (INC) algorithm [12], which looks at incremental and to categorise the discussed 50 MPPT methods into eight categories
momentary conductance of PV systems, can track the maximum based on their tracking nature and manipulation techniques for
power point (MPP) of a PV system and exchange high PV power finding the true MPP. The categorised eight groups are
to the load. The research work presented in [13] clarifies the conventional methods, methods based on mathematical
misconception between the widely used P&O and INC algorithms calculations, constant parameters based methods, measurement and
and shows that they are almost highly identical under steady-state comparison based methods, trial and error based methods,
and transient conditions. It is shown that they both have similar numerical methods, intelligent prediction based methods, and
mathematical expression except that the INC ignores the higher- methods based on iterative in nature. To find out the efficient
G
Iph = [ISCC − STC + Ki(T − T STC)] , (2)
GSTC
where V OCV − STC (in volt, V) is the open circuit voltage (OCV) at
Fig. 2 I − V and P − V curve of a PV cell [2] STC, Kv is the OCV coefficient, V th refers to the thermal voltage of
the cell, these values are available on the data sheet provided by
method among others, a tabular comparison is also done in each module's manufacturer. With V PV and simplified IPV, the power
category based on 11 selection parameters. The considered 11 produced by the PV module is represented as [24]
selection parameters are design complexity, sense parameters, PV
array dependency, prior training, periodic tuning, convergence qV PV V PV
speed, analogue/digital in nature, cost, tracking true MPP, stability, PPV = V PV × Np Iph − Ioexp − . (4)
Ns AKT Ns
and efficiency of the system. These comparison tables and ways of
categorisation will be helpful in future for selecting appropriate
MPPT methods for the solar PV system. The I − V and P − V characteristics curve of the solar cell is shown
The rest of the paper is organised as follows. In Section 2, a in Fig. 2 [2]. The curve in this figure indicates that the operating
model of the solar PV system with its I − V characteristics, point of the PV does not remain at a stable point; it actually varies
equivalent circuit, effect of temperature, insolation, and PSC on from zero to open-circuit voltage. There is only one point, which
maximum power are presented. The need for the MPPT controller, enables maximum power for a given set of solar insolation and
its selection parameters, and PSC supported MPPT methods are temperature level. That particular point is indicated as MPP and at
described in Section 3. Overview of various MPPT tracking that point, the current and voltage that are found are presented as
techniques and their comparisons are demonstrated in Section 4. IMPP and V MPP in Fig. 2. This significant MPP and its numerous
An overall discussion about the discussed methods after reviewing tracking methods are the main focal point of this review work.
them is also detailed in Section 4. Finally, the paper concludes in
Section 5. 2.2 Effect of irradiance and temperature
The output of PV shifts with the changing climatic conditions [27,
2 Model of PV cell 28]. Since the irradiance of the solar cell relies upon the incidence
The model of a solar PV cell is an important part of analysing a PV angle of the sunbeams, this parameter straightforwardly influences
system. Its modelling is classified into three sections, which are the output adjusting the P − V and I − V characteristics [20]. The
described below. output current, IPV, of a PV module is broadly impacted by a
variety of sun-oriented irradiance, G, though the output voltage,
2.1 I − V characteristics and equivalent circuit V PV, remains practically constant. On the other hand, for a
changing temperature, it is found that the voltage shifts generally
PV begins from two separate words – photo, which implies light, while the current remains practically unaltered [9].
and voltaic, which alludes to the production of power [22]. The temperature of the PV cell increases because of three
Subsequently, the term PV brings the significance of producing reasons: (i) its own heat amid PV activity, (ii) the energy emanated
power specifically from the sun. A sun-powered array comprised at the infrared wavelength, which has a warming impact on the
several combinations of sun-based modules, where every module cell, and (iii) an increase in sunlight-based insolation [10]. The
comprised various solar cells [1]. Solar cells comprise p–n diodes OCV, V OC, and SCC, ISC, are found by (5) and (6), respectively, of
manufactured in a thin layer of semiconductor [23]. They resemble PV array at sun intensity and cell temperature, where the sun
p–n diodes and their attributes are additionally comparative. Fig. 1 intensity (insolation) is measured in W/m2
displays the equivalent circuit of a perfect PV cell [1]. This ideal
structure is sufficiently precise to comprehend the PV attributes ∗
V OC = V OC + a2(T − T ∗) − (ISC − ISC
∗
)Rs, (5)
and the reliance of the PV cell on varying climatic conditions [20].
Alongside temperature and insolation, the partial shading also 3.3 MPPT controller under PSC
demonstrates a noteworthy impact on solar cell characteristics [32,
33]. The circumstance in which diverse parts of a PV module get a Some well-recognised MPPT methods under uniform irradiance
distinctive measure of irradiation is alluded to as PSC. Shading and PSC are depicted in Table 3. The comparison among the PSC
might be caused by mists, contiguous structures, and trees [34]. supported MPPT methods is shown in Table 4. The methods are
Along these lines, the most imperative and demanding stint for the compared in terms of six selection parameters, such as PV array
MPPT controller is achieving MPP under PSC. In a PV array, when dependency, convergence speed, periodic tuning, complexity,
a cell is shaded and gets low insolation or no insolation, according analogue/digital in nature, and sensitivity. From Table 4 [38, 43–
to (2), the photocurrent of that cell (Iph) reduces. Since, in view of 45], it is seen that most of the methods under PSC are digital in
Kirchoff current law, the current of all series connected PV cells nature and highly complex. Moreover, in terms of convergence
must be the same, the inside diode of the shaded cell goes into the speed, periodic tuning and PV array dependency, the PSO, genetic
breakdown region to repay the lessening of photocurrent. Thus, the algorithm (GA), differential evolution (DE), biological swarm
shaded PV cell acts as a load rather than a generator [34]. chasing (BSC), PV output senseless (POS), SC, firefly algorithm
(FA) and ant colony optimisation (ACO) performs better under
PSC than the other classical methods. However, as the
3 MPPT controller characteristics of the PV array under PSC comprises many local
The system that operates the PV in such a way to extract maximum MPPs and one global, the classical MPPT algorithms cannot
power is termed as the MPPT controller. If the controller works accurately track the GMPP. Therefore, the GMPPT algorithm is
deliberately at MPP, independent of the climatic condition, the needed [46]. In the literature, the GMPPT algorithm is classified
efficiency of the PV system is enhanced. This should be possible into two classes [47, 48]: (a) firmware-based and (b) hardware
by legitimately coordinating PV source with the load for any architecture-based algorithms. The firmware-based GMPPT
climate condition to accomplish maximum power production. methods use the hybrid approach by combining two MPPT
There are two processes by which maximum power can be algorithms. Firstly, the approximate GMPP is located using search
extricated from the PV array and they are: mechanical and algorithms and then, it is accurately tracked using classical MPPT
electrical tracking. In mechanical tracking, the PV panel direction algorithms. The hardware architecture-based algorithms are related
changes according to the changes of months and seasons to the power converter topology and design of the PV system. Soft
throughout the year, while in electrical tracking, the I − V curve is computing techniques such as PSO, grey wolf optimisation
used for locating MPP [10, 35, 36]. MPPT is an integral (GWO), and CuS optimisation can tackle the GMPP of a PV
component of modern power systems, which ensures the system under the same PSC [16]. However, these techniques suffer
penetration of maximum power to the load/batteries/motors and the from high oscillations around the GMPP and they cannot
power grid, for off-grid and on-grid applications, respectively. dynamically track the new GMPP when it changes its position. To
Since the conversion rate of sun energy to electrical energy of PV solve these problems, a hybrid GWO–fuzzy logic control (FLC)
arrays is still low and the solar irradiance is not always uniform, algorithm is proposed in [49], which is able to track dynamic
the MPPT controller finds its widespread application in PV plants. GMPP. The global perturbed-based extremum seeking (GPESC)
A brief discussion on the necessity of the MPPT controller is scheme is proposed in [48] and the ability of this scheme to track
presented in Section 3.1. the GMPP is analysed for both static and dynamic PV patterns
which shows a high value of tracking accuracy (99.99%).
3.1 Need for an MPPT controller
4 Review of various MPPT algorithms
Any environmental changes impose imperatives on power
production from a sustainable power source. Especially, the impact Owing to its versatile use, researchers attempted a lot to soak up
is more serious in solar and wind energy systems. Additionally, the maximum amount of power from the panel. Until now, a lot of
wind and solar systems confront challenges on (i) changing MPPT methods have been developed. Each technique has its own
climatic conditions and (ii) grid incorporation [2, 37]. Henceforth, types of operation processes, advantages, disadvantages, and
solar PV and wind energy conversion systems embrace MPPT applications. To classify the available methods there is no proper
procedures to give supportable power output [38, 39]. For this assessment since one might be useful for a particular application
reason, it is necessary to ensure that there exists an MPP in I − V and not for other, again one can be extremely efficient but another
and P − V curve for variable irradiation and temperature. This is not. In this review work, the discussed 50 methods that are
MPP continuously moves its position when any environmental classified into eight groups based on their tracking nature are
change happens. Therefore, MPPT controllers are designed to shown in Fig. 3.
continue the tracking of MPP and they form an indispensable piece
of the PV system. The presence of a controller adequately modifies 4.1 Conventional methods
the resistance seen by the panel and consequently, urges the panel
to work nearer to MPP [37]. Efficient MPPT controllers are The P&O, INC, and HC methods and their modifications are
essential to modify the operating point of the load associated with classified as conventional methods as these methods have existed
changing the duty cycle of the converter. for decades. The details of these conventional methods are listed in
the following sections.
Table 2 Comparison among commonly used analogue/digital MPPT ICs or micro-controllers [41]
ICs/microcontroller Tracking speed Used in MPPT methods
FPGA XC2C384 85 ms P&O
microcontroller Intel 87196 500 ms INC
analogue circuitry 20 ms ESM
DSP TMSF28335 80 ms load-current based MPPT
DSP TMS320F28335 1.5 s FLC-based P&O
FPGA Virtex II 30 ms ANFIS
DSP TMS320F240 20 ms P&O-based PI
PC PENTIUM IV 2.4 GHz 16 ms INC
Table 3 Some well-recognised MPPT methods under PSC and uniform irradiance condition [1, 10, 38]
Uniform irradiance condition PSC
i. P&O i. partial swarm optimisation (PSO) xi. Beta
ii. HC ii. GA xii. SC
iii. INC iii. DE xiii. Cuckoo search (CuS)
iv. OCV iv. ANN xiv. FA
v. short circuit current (SCC) v. FLC xv. ABC
vi. RCC vi. BSC xvi. ACO
vii. current sweep vii. POS xvii. COS
viii. DC link capacitor droop control viii. state based (SB) —
ix. load I or V maximisation ix. array reconfiguration (AR) —
x. dP/dV or dP/dI feedback control x. PC —
4.1.1 P&O method: This strategy is one of the least complex derived the real PV power by observing the variations in power
strategies, which has been considered by various scientists [50–52]. [34]. Until reaching the MPP, this situation is recapitulated. The
In this procedure, the voltage of the PV array is perturbed and MPP is achieved when the changes of power with respect to
the adjustment in its output power is noticed. At each cycle, the changes in voltage being zero (dP/dV = 0). The essential
voltage and current of the PV are measured by the tracker and flowchart of the P&O calculation is shown in Fig. 4 [9]. The figure
shows that the algorithm starts by obtaining the present value of V this type of method. Various types of modified P&O methods are
and I at V(k) and I(k) and the power at P(k). The algorithm found in the literature [56–62]. The most common ones can be
perturbs the operating voltage, V, at every MPPT cycle with categorised as below.
respect to the reference voltage V ref . The operating voltage, V,
oscillates until the MPP is reached to provide the ideal operating i. Conventional P&O
voltage V MPP. Here, CP indicates the losses of power due to the step
width of the perturbation. The most important focal point of this • fixed perturbation
technique is that the information of the solar cell qualities is not • adaptive perturbation
needed and can be connected to any PV system. Notwithstanding, ii. Modified P&O
this technique has a few downsides, as detailed in [53–55].
• fixed perturbation
4.1.2 Modified P&O method: In the modified P&O strategy, the • adaptive perturbation
perturbing element is modified in order to achieve the desired
output. This method has some circuit complexity but the tracking A comparison of the modified P&O methods with the conventional
speed of this method is higher and exact compared to the P&O P&O method has been done and depicted in Table 5 which
algorithm. The estimated perturb–perturb method is one kind of indicates the variation among those methods.
4.1.3 INC method: The algorithm of INC MPPT depends on progression esteem is automatically adjusted relying upon the
dP/dV, which is equivalent to zero for the purpose of the MPP. It separation between the MPP and the working point of it. This
was proposed to enhance the tracking precision and dynamic strategy presents an optimising method with a low oscillation.
execution under quickly changing conditions [63–65]. The
flowchart of the INC algorithm is shown in Fig. 5 [44, 66]. The 4.1.5 HC method: The HC operation is fundamentally the same
algorithm begins with the cycle by finding the value of V(t) and as P&O; the main distinction is, rather than iterating the voltage or
I(t) at time t. Comparing the instantaneous conductance with INC, current, it updates the working purpose of the PV exhibited by
MPP is tracked. Based on the comparison of the output, the control perturbing the duty cycle, d(k), by a settled step size, towards
voltage (V ref ) is adjusted. The corresponding equation involved in rising power. Since the duty cycle is changed specifically without a
this method is detailed in [67–70]. The INC offers the benefit of proportional–integral (PI) controller, the HC strategy is in some
better execution under quickly changing climatic conditions and cases known as the direct duty cycle procedure [71]. The flowchart
additionally provides less oscillation around the MPP. Therefore, of this technique is shown in Fig. 6 [38], where sense array's
the efficiency of INC to accomplish MPP is about the same as the voltage, current, and power are indicated by V(k), I(k), and P(k),
P&O technique [68]. The disadvantage of this technique is, under respectively. The significant downside of this technique is that
PSC, it cannot track the true MPP. The control hardware of INC is increased oscillations are seen close to the ideal working point
complex which prompts a high system cost [69, 70]. compared to the maximum extractable power, i.e. it neglects to
meet at MPP, and the time taken to achieve steady-state relies upon
4.1.4 Modified INC method: The rigid step-size is one of the the estimation of the initial duty cycle and step size. In some other
downsides of the INC strategy [21]. When this progression is research works, versatile renditions of HC have been created [34,
picked vast, the tracking rate turns out to be quicker, however, adds 60].
a significant amount of steady-state oscillations in power. When
the step-size is small, oscillations diminished; however, the 4.1.6 Modified HC method: This method is a modified version of
tracking procedure turns out to be slower. The adjusted INC the HC method. In [60], Lei et al. proposed a duty cycle clear
strategy comprises changing the rigid step-size in which the strategy in view of the modified HC to procure for the PSC. In this
4.2 Methods based on mathematical calculations 4.2.5 Analytical solution (AS) based method: This technique
In this category, the methods described below use various depends on the mean value theorem. It provides an AS for
mathematical calculations based on their respective equations to approximating the position of the MPP. In this method, the current,
locate the MPP. From the existing methods of the MPPT controller, the voltage of the PV system and the duty ratio of the controller are
eight methods are listed in this category, and their details are used as the independent variables and the output power is used as a
enlisted below. dependent variable. This method is also entitled as a bracketed
method.
4.2.1 Incremental resistance (INR) method: One of the notable
algorithms created to track the MPP of the PV power system is the 4.2.6 Curve-fitting-based MPPT: The qualities of the PV panel
INR. The main thought of the INR-based tracker is that PV power and its manufacturing details are required in these strategies for
derivative w.r.t its current is zero at the MPP. The mathematical finding the true MPP. It is a mathematical-based procedure. The
model for extracting the PV power by this method is detailed in preferred standpoint of this technique is its effortlessness, as no
[16, 72, 73]. differentiations are to be ascertained. However, the main
impediment to this strategy is that it needs a huge memory for
computing the mathematical formulae included.
4.2.2 Variable step-size INR method: To enhance the accuracy
and speed of tracking, the INR method is a powerful approach but
it has some drawbacks. The moderate speed and oscillations near 4.2.7 Linear reoriented coordinates method (LRCM): The
the MPP are the drawbacks of the INR method, but these MPP equation is calculated iteratively in this method, where the
drawbacks can be solved in the variable step estimate INR strategy equation is controlled to locate a surmised representative for the
[73]. In this method, when the system is far from the MPP, the step MPP. The preferred standpoint of this strategy is that an estimated
measure turns out to be little as the systems approach the MPP, representative solution will be found for exponential functions
which diminishes oscillations around the MPP and provides without symbolic solutions.
quicker speed.
4.2.8 Ripple correlation control (RCC) method: The conversion proportional to its OCV, V OC. This technique applies a nearly linear
system's ripple is considered in this method to determine the MPP relationship between the OCV and the maximum output voltage,
of the system [76]. The current, voltage and power ripples are V MPP, of the PV array under changing barometrical conditions.
caused by the switching of the power converters. These ripples are This technique can be approximated by [1]
utilised here to track the MPP. The ripples (high frequency) in
power and voltage are occupied utilising channels which are then V MPP = KOCV OC (KOC < 1) . (8)
used to find dPPV /dV PV [77]. The variety dPPV /dV PV is then used
to track MPP in this method. This method is also named as where KOC is a constant that shifts in the vicinity of 0.7–0.8 relying
‘physics-based strategy’ [78]. upon the PV cell trademark. Generally, it is experimentally
A tabular comparison based on the selection parameters has measured in view of V MPP and V OC after deciding the estimation of
been made among different mathematical calculation-based
KOC under changing barometrical condition [34, 67]. A flowchart
methods and is shown in Table 7. The comparison table indicates
that most of the methods of this category are expensive, stable and outlining this technique is shown in Fig. 7a [21].
digital in nature. In terms of convergence speed, the curve fitting,
INR, variable step size (VSS) INR, β, AS and RCC shows faster 4.3.3 Fractional SCC method: This technique uses another
response than others. constant parameter approach [79–82], which is generally like the
OCV strategy (as shown in Fig. 7b [21]). There appears an
4.3 Methods based on constant parameters additional linear relationship between the SCC (ISC) of the sun-
powered panel and the MPP current (IMPP), which can be depicted
In this category, the MPPT methods utilise a predefined fixed value by the accompanying equation:
in order to track the MPP required. About 15 methods are listed in
this category and they are delineated below. IMPP ≃ K2ISC . (9)
4.3.1 OCV pilot PV cell method: In the pilot cell MPPT Similar to the OCV technique, the load ought to be shed in order to
technique, the constant voltage or current technique is utilised. This decide the ISC. The SCC strategy is more precise and proficient
strategy utilises just a single feedback-loop control and stays away than the OCV technique [67, 82] because of pragmatic issues
from the issues produced by the interferences of the working of PV related to measuring ISC but its usage costs are significantly higher.
in the past techniques. Then again, this strategy experiences the
error of reading the cell voltage because of PSC, yield confounds
between the MPP of the PV generator and the pilot cell. 4.3.4 p–n junction drop voltage tracking method: It is known
that the temperature characteristic of the p–n junction diode is
similar to that of the solar cell. For this reason, at the posterior of
4.3.2 Fractional OCV: The OCV [67] MPPT algorithm is an the solar panel, diodes are introduced in order to recognise the
improved strategy for an off-grid application. This strategy is changes of surface temperature by a drop of the voltage of the p–n
feasible as the output voltage of the PV at MPP is linearly junction [21, 83]. This concept is utilised here. This method is also
known as ‘excellent operating point tracker’. This strategy 4.3.8 Temperature parametric (TP) method: Similar
considers just the impact of temperature and ignores the impact of researchers [85] of the past strategy also demonstrate the optimal
insolation, which prompts a non-precise tracking method [21]. voltage, V MPP, utilising trial tests and define it as
4.3.5 Constant voltage (CV) method: The CV technique is the V MPP ≅ (u + Sv) − T(w + Sy) (10)
least difficult MPPT method [21]. It depends on the perception that
the MPP happens in the vicinity of 72–78% of the OCV, V OC, for where u, v, w and y are the PV parameters under various irradiance,
the standard climatic condition [32]. This method overlooks the S. This technique needs the estimation of temperature and
impacts of insolation and temperature and accepts that the steady insolation to locate the optimal voltage.
reference voltage is an adequate estimate of the genuine MPP. Only In this category, eight MPPT methods are detailed briefly.
a voltage sensor is required in this method to gauge the array Based on the selection parameters of these methods, a tabular
voltage V PV to set up the duty cycle of the DC/DC converter. comparison has been made as shown in Table 8. The table indicates
that most of the methods of this category are not stable and only
4.3.6 Constant current (CC) method: The CC strategy depends applicable for off-grid applications. Prior training is necessary for
on a similar marvel of the CV technique. In the CV technique, the most of the methods of this category. If the tracking accuracy of
PV array works at the constant voltage and in this strategy, PV these methods is considered, the OCV pilot cell, TP, and BFV
array works at the steady current. The MPP touches base in the methods have higher tracking accuracy than others.
vicinity of 78 and 92% of the SCC, ISC. Therefore, the detected
parameter is SCC in this method. 4.4 Intelligent prediction based methods
These methods are mainly soft-computing-based techniques, which
4.3.7 Best fixed voltage (BFV) method: A pre-determined are now becoming a powerful tool to accord with MPPT
period is intended for information gathering about insolation and optimisation. Besides, the accessibility of the elite and reasonable
temperature, is utilised by BFV to estimate the MPP [84]. The micro-controller makes the usage of these calculations conceivable
controller sets either the PV array's operating point to the BFV or under practical conditions. The FLC, artificial neural network
the output voltage to the nominal load voltage. The description and (ANN), evolutionary algorithm (EA), and hybrid adaptive neuro-
flowchart related to this method are detailed in [84]. fuzzy inference system (ANFIS) methods are listed in this category
and they are described below.
4.5.2 ACO algorithm: The quest for searching the path for food
by an ant is the basic concept of this method.
In this method, at first, an individual ant chases the way
arbitrarily and after that the rest of the ant inputs pheromone to
take after, shaping positive feedback as introduced and clarified in
[97]. The thickness of the pheromone depends on the amount of
tracking of a specific path. The nature of finding the shortest and
random path in the ACO method is shown in Fig. 11a [20, 98].
Different usages of the ACO for the MPPT have been introduced in
[99–101].
4.5.7 Artificial bee colony (ABC) method: The inspiration for 4.6.2 Load current or voltage maximisation (LCVM)
the ABC method comes from the model displayed in [20]. It method: This MPPT method is developed by assuming the PV
delineates the rummaging example of the honey bee colonies. This power converter as a lossless converter. Since the converter is
procedure is generally a meta-heuristic plan based on the swarming lossless, when a PV module is connected to it, the PV power
technique. It was first presented for taking care of multi-modal and maximises the output load power, conversely, the output load
multi-dimensional optimisation issues. Again, in three categories power maximises the PV power. Loads that are used in the PV
the honey bees are sorted into onlooker, scouts, and employed. In system are the mostly resistive type, voltage-source or current-
this method, a candid arrangement of the improvement issue source type. Therefore, this method utilises the process of
indicates the area of a sustenance source and the nature of the maximising the load voltage or current to maximise PV power.
related solution is indicated by the food source. This method suffers a lot for the accuracy of tracking the proper
The methods based on iterative in nature are mainly digital in MPP as it considers a lossless converter.
nature, can predict true MPPT and is not dependent on PV array.
These methods are expensive and only applicable to off-grid 4.6.3 Look-up table method: In this technique, prior information
applications. Based on the selection parameters, a comparison is of the PV panel material, technical information, panel attributes
shown in Table 10, which helps to find the best among them. under various natural conditions is required and stored in order to
locate the MPP. A new voltage V MPP is dictated at each cycle by
4.6 Measurement and comparison based methods the controller in view of the correlation of the deliberate
estimations of temperature and insolation with the information
These methods are established on detecting the external parameters
stored in the look-up table. The look-up table is created in light of
such as PV current, voltage, insolation or temperature and
the specifications provided by the manufacturer or by means of
contrasting them with a pre-defined MPP for determining the
exploratory tests on the PV under various environmental
required MPP. The details for the included methods in this category
conditions. This method has the downside of expansive memory
are represented below.
necessities for look-up table storage.
4.6.1 Current sweep method: The current sweep waveform of
4.6.4 Linear current control (LCC) method: This strategy
the PV array is utilised in this method to find the required I − V
exploits the linear relationship amongst the V and I of the
curve. To achieve the voltage at MPP, the sweep is repeated
characteristic curve. In this method, to assess the MPP, the power
continuously until it reaches to the desired MPP. Although this
and current equations are linearised through algebraic equations
strategy is complicated, slower in speed and losses of power occur
[109, 110]. For linearisation of this technique, higher insolation
during the sweeping, it is efficient in tracking the real MPP [10].
conditions are needed. Here, the technique graphically recognises strategy is if the tangent is parallel to the x-axis then this technique
the converging purpose of two curves, the power curve of PV will not work at any stretch of the imagination.
indicated by f (P, I) = 0 and maximum output power curve
indicated by (dP/dI) = 0. 4.7.5 Secant method (SM): SM is a straightforward and
In this category, four MPPT methods have been discussed. extremely efficient method for finding the MPP [121, 122]. It is
Among these methods, only the LCC method is expensive. In order exceptionally productive, does not require sectioned points of
to select the best method among this category, a comparison in confinement and even requires only two initial guesses.
tabular form has been made in Table 11. The table provides a better Nevertheless, it has some drawbacks that if the initial value is not
explanation for finding the best MPPT methods among this near the MPP, at that point strength and merging are not ensured.
category for a specified application.
4.7.6 Bisection search method (BSM): This strategy is a
4.7 Methods based on numerical methods straightforward one, as it requires no derivative estimations,
economically cheap in execution as the number of sensors required
The methods included in this category use various numerical
are less and the technique is quick in MPP tracking. Research
methods in order to find the peak of the MPP. Newton–Raphson
works in [123] presents a reasonable case of using the BSM, while
(NR), secant method (SM), false position method (FPM), bisection
[124] provides an adjusted method of it. The BSM is a better
search method (BSM), and Brent method (BM) are included in this
numerical technique for discovering roots of conditions in order to
category. The details of these methods are presented below.
track the true MPP [19].
A tabular comparison among seven numerical methods based
4.7.1 Brent method (BM): This method is a standout amongst the MPPT has been made in Table 12 on the basis of various selection
most productive strategies to track the MPP, introduced in [111] standards. The table presents valuable information about this
and clarified in [20, 112, 113]. Moreover, it gives quick merging to category that these methods may be expensive in cost but have
the MPP using the SM and ensures meeting as a result through the high accuracy and fast convergence speed.
BSM. A resilience esteem and three voltage points are required for
locating the MPP in this method. Among SM, inverse quadratic
4.8 Methods based on trial and error
interpolation or BM, the best-ascertained arrangement is given by
the BM. These techniques utilise a trial of estimation and observe the
subsequent outcome, which decides the bearing criteria for the next
4.7.2 Central point iterative (CPI) method: CPI technique endeavour in order to find the peak of MPP. The methods included
partitions the whole parameters of the controller into four non- in this criterion are presented below.
covered interims giving three centre points. PV voltages of these
three centre points are computed and then utilised for locating the 4.8.1 DC-link capacitor droop control: This technique depends
MPP. Therefore, two out of the four non-covered interims are on a standard of energy equilibrium at the DC link [40]. In this
chosen and the other two are rejected. The chosen two interims are structure, the boost chopper keeps the line voltage to be consistent
again split into four non-covered interims, for next power by a feed-forward loop and the pulse-width modulation (PWM)
emphasis. Thus, this method works. inverter utilising a feedback loop, controls the load voltage.
Fig. 12a [21] indicates the control conspire for DC link capacitor
4.7.3 False position method (FPM): Owing to the iterative droop control. The DC-link capacitor (Cdc) decreases the voltage
formula of the FPM, it is also familiarised as the Regula Falsi ripple of the DC–AC inverter and furthermore gives a hold-up time
method, which is clarified in [114, 115], while a case of the amid which the insolation swings immediately amongst high and
modified FPM has been displayed in [116]. Although it meets the low.
root quicker than BSM, it is slower than the NR approach.
4.8.2 POS control method: This procedure is devoted to that PV
4.7.4 NR method: The strategy is usually utilised due to its system, where the system of tracking speculates only by the
effortlessness and incredible speed. This scheme has been utilised flowing of current to the load (as shown in Fig. 12b [21]). A PWM
by numerous researchers to find the parameters for tracking the signal is utilised for the power transformation of the system and the
MPP, a portion of the significant distributions is detailed in [117, associated control circuit is straightforward. For an extensive PV
118]. Research works in [119] provide enhanced MPPT control system, this method can be worked adequately and more securely,
through the modelling of the PV array. Moreover, Pradhan and since the feedback V and I of PV modules are not required [21].
Subudhi [120] give an auto-tuning-based adaptive MPPT for a PV
framework, which utilises the NR technique. This strategy gives 4.8.3 Three point comparison method: As the name
quick convergence and great precision. The downside of this demonstrates, the three-point technique looks at power at three
focuses on the P − V curve for MPP. This strategy dodges the
oscillation issue of the P&O algorithm by two focuses; the past and
the present focuses. The technique of this method runs
intermittently by perturbing the terminal PV voltage and looking at
the output power of the PV for three purposes. They are [21]: (i)
the current operation point A, (ii) a point, B, perturbed from point
A and (iii) a point, C, doubly perturbed in the opposite direction
from point B.
Consequently, for locating the MPP, nine conceivable cases
must be executed in the control circuit. These cases are delineated
in Fig. 13 [21]. The working procedure according to the nine
conceivable cases of power variation by the three-point weight
comparison algorithm and its flow diagram is detailed in [125].
comprises a DC/DC boost converter to exchange energy to load A tabular comparison among trial and error based methods has
from the PV panel and to track MPP by the working panel at V MPP been made on the basis of various selection standards and is shown
(voltage at MPP). The hardware implementation of this method is in Table 14. The table acts as a significant tool for finding the best
shown in Fig. 15 [10]. Different kinds of DC/DC converters, such MPPT methods in this category for a specified application. A
as boost, buck–boost, flyback, zeta, cuk, single ended primary discussion of the whole methods detailed in this paper is described
inductor converter (SEPIC), coupled inductor combined cuk in the next section.
SEPIC (CI-CCS), are adopted to fulfil the purpose of MPPT of the
solar PV system. Therefore, the comparison based on input current, 4.9 Discussion on findings
computation performance, output to input voltage ratio and nature
of output voltage, among various kinds of DC/DC MPPT Researchers and scientists around the world have been working for
converters are presented in Table 13 [129–131]. The output of the years to improve the efficiency of the solar panel and the MPPT
converters depends on the duty cycle. Among different types of has been a centre of interest for many researchers. A number of
converters, the CI-CCS provides a bipolar output voltage. For an new MPPT algorithms have been already developed to extract
on-grid PV inverter, an efficient control method is proposed in maximum power from the PV array. Choosing a specific MPPT
[132] based on the ANN-MPPT in conjunction with an SC to avoid method for a specified application among the existing methods is a
the utilisation of the DC/DC converter with two controllers. difficult task as each method has its own pros and cons. In this
However, the downsides of the SC method are that the output review work, a comprehensive review of 50 MPPT methods are
voltage and power vary with the increase in exchanging the power presented from different works of literature along with their merits
between the grid and PV system [14, 127, 128]. and demerits. After reviewing the research works on the discussed
MPPT methods of the solar PV system, the following discussion
can be drawn, which may be a useful tool for selecting the most
4.8.7 Extremum seeking method (ESM): This method is based
constructive and appropriate type of MPPT to fulfil the
on the extreme seeking control theory, which provides an
requirements of both operators and consumers.
oscillatory behaviour around the equilibrium position. The
estimated position of the MPP is not needed in this method for
• Conventional methods such as P&O, INC, and HC are simple
locating the actual MPP. This technique provides an efficient
since their circuit of implementation is not complex, but they are
extraction of solar energy as well as ensures the stability of the
less accurate and slow in response and present a problem of
system [118].