School of Electronic and

Electrical Engineering
FACULTY OF ENGINEERING

ELEC3875

Project Report

Project Title

Design and MATLAB/Simulink simulation of Hybrid Maximum Power Point Tracking Control

Name

Fahad Sandel

SID: 201310624 Project No. 24

Supervisor: Dr. Salma Alarefi Assessor: Dr. Amir Abiri

1
 School of Electronic and
Electrical Engineering
FACULTY OF ENGINEERING

ELEC3875 Individual Engineering project

Declaration of Academic Integrity

Plagiarism in University Assessments

and the Presentation of Fraudulent or Fabricated Coursework

Plagiarism is defined as presenting someone else’s work as your own. Work means any
intellectual output, and typically includes text, data, images, sound or performance.

Fraudulent or fabricated coursework is defined as work, particularly reports of laboratory

or practical work that is untrue and/or made up, submitted to satisfy the requirements of a
University assessment, in whole or in part.

Declaration:
• I have read the University Regulations on Plagiarism [1] and state that the work
covered by this declaration is my own and does not contain any unacknowledged
work from other sources.
• I confirm my consent to the University copying and distributing any or all of my work
in any form and using third parties (who may be based outside the EU/EEA) to
monitor breaches of regulations, to verify whether my work contains plagiarised
material, and for quality assurance purposes.
• I confirm that details of any mitigating circumstances or other matters which might
have affected my performance and which I wish to bring to the attention of the
examiners, have been submitted to the Student Support Office.

[1] Available on the School Student Intranet

Student name: Fahad Sandel Project No. 24

Signed: Fahad Date: 12/05/2023

2
Abstract

As time goes on, there is a growing need for power. RES are also becoming more popular
since they are a noiseless, clean, more affordable, and limitless source of energy. Since
photovoltaic energy is anticipated to overtake other renewable sources in the next years,
the solar PV industry has seen tremendous growth. The difficulty of maximizing efficiency
under various irradiance conditions, on the other hand, is one that photovoltaic systems
must overcome. A PV system's I-V and P-V characteristic curves are essentially nonlinear
because of the sun's shifting behavior, manufacturing tolerances, ageing, shadowing, bird
droppings, and the sun's beams' different incidences on different solar cells. Partial
shading is the cause of how the PV system behaves. Consequently, maximum power point
tracking (MPPT) has become a major problem for PV systems.

3
Acknowledgements

I would like to dedicate this segment to thank and acknowledge all the people that helped me
see this project to its end.
I would like to thank Dr. Sadegh for providing us the lectures and materials needed for this
project.
I would like to thank my advisor, Dr. Salma Alarefi, who advised me and helped shape this
project to what it is with all the feedback I got.
I would like to thank my family who encouraged me and pushed me to do my best throughout all
my years abroad.
Finally, I would like to thank my friends who were with me through the dark times and the bright
time.

4
Table of Contents:

1- Introduction ……………………………………………………………………………… 1
2- Background ……………………………………………………………………………… 2
3- Literature Review ………………………………………………………………………. 3
3.1- Description of the Partially Shaded PV Energy System ………………………. 4
3.1.1- Description of the Power and Control Circuits ……………………………. 5
3.1.2- Description of UI, PS, and CPS under Study ……………………………… 6
3.2- Previous MPPT Techniques for PV Energy System ……………………………7
3.2.1 Incremental Conductance (InC) Technique ……………………………………. 8
3.2.2 Dragon Fly Optimization (DFO) Technique …………………………………… 9
4- Project (Proposed Hybrid (InC-DFO) Technique for PV Energy Systems) ……… 10
4.1- Design ……………………………………………………………………………11
4.2- Simulation Model ………………………………………………………………………12
4.3- Result & Analysis …………………………………………………………………13
4.3.1- Case 1. Uniform Irradiance Condition……………………………………… 14
4.3.2- Case 2. Partial Shading Condition …………………………………………15
4.3.3- Case 3. Partial Shading Condition …………………………………………16
4.3.4- Case 4. Complex Partial Shading Condition ………………………………17
4.3.5- Case 5. Field Atmospheric Data of Beijing ……………………………… 18

5- Conclusion ……………………………………………………………………………………… 19
6- References ……………………………………………………………………………………… 20
7- Appendix ……………………………………………………………………………………… 21

5
 1. Introduction

Electrical power is increasingly in demand. As a result, the consumption of fossil fuels has
increased, which has clear detrimental effects on the environment, including the production of
carbon dioxide and Sulphur dioxide, which contribute to the phenomena known as the greenhouse
effect. RES are currently replacing traditional energy sources like fossil fuels, and their role is vital.
As a result, many countries, including Iceland, Sweden, the United States, Costa Rica, and
Nicaragua, are switching to RES. Since the turn of the century, among all RES, solar radiation has
become a promising source of power and gained more attention due to its status as a silent, A
pollution-free, steadily decreasing in price, and virtually limitless supply of power. PV energy had
also brought strong growth to the solar PV market, and based on current developments in solar
technologies, it is expected to surpass other RES too though.

The challenge of maximizing efficiency in the face of extremely fluctuating irradiance

circumstances has been one of several that PV systems have faced in the meanwhile. A
photovoltaic system's I-V and P-V characteristic curving hips are intrinsically nonlinear because of
things like the sun's unpredictability, manufacturing tolerance, ageing, shadowing, birds trying to
drop solar cells, and solar cells' various orientations towards the sun [1]. The PV system's non-
linear behavior is brought on by partial shading [2]. Therefore, the most urgent problem with PV
systems is output power (MPPT).
B

The PV system has been subjected to multiple voltage level (MPP) approaches, such as
perturb and observe (P&O) and perturb and observe (Inc.). The detection of GM also employs P-V
and I-V curves. The techniques in [4] are graded based on how fast they perform tasks, how exactly
they create outcomes, and how challenging it is to put them into practice. Fractional Bipolar
Transistor Current (FSCC), particle swarm optimization, and face breaker are among the
conventional and often employed techniques in PV energy technologies [5].
C

6
 2. Background:

The FOCV approach requires only a single additional feedback loop, making it easy to
implement. InC can employ analogue devices rather than microcontrollers or other digital
components. To keep an exact open-circuit voltage, PV module performance must be continuously
opened [6]. The efficiency of this automated system is low because power loss occurs during the
measurement of Voc rather than the predicted inaccuracy in KOC. Because the correlation between
Voc and Vmpp must be approximated, this method cannot choose precise MPP values.
C
However, unlike the FOCV method, obtaining FSCC requires an extra switch and a brief
period of shorting the PV array. Similar to the FOCV method, one such approach relies on
approximations that prevent it from following true MPP. It is proposed that the pv systems voltage
sweeping process be used to overcome this obstacle and achieve true MPP. Because of this, the
system's power cost will rise [7]. The algorithm needs to be calibrated, so the V1 is set to 0.05 Voc
or the V2 is set to 0.95 Voc, making the MPP between 5% and 95% of Voc, within which Voc seems
to be the transparent voltage under conventional controlled circumstances (STC) [8]. D
P&O differs from traditional MPPT techniques in that it involves tampering with voltages and
then monitoring the system's output power. To find the MPP, the homogeneous power perturbation
process is repeated if the old power is smaller than the new power. If you don't follow MPP, the
perturbation process will be reversed. Changing weather conditions are a problem for P&O because
of the energy lost due to the system's oscillation around MPP. These complications can be
remedied by employing a reference voltage with a finely tuned duty cycle and discrete step size.

Adaptive P&O techniques, which vary the size of each step, are proposed by the authors [9] as
a means of mitigating P&O's drawbacks. The DC-to-DC converter's step size can be stabilized by
adjusting the pulse signal technique used depending on the location of the switching point. A larger
number of iterations is used when the MPP is located far from the operating point. When the pulse
signal is closer to MPP, its size decreases. Since this is possible at MPP, the step size can be
reduced to zero. E
By fusing the customized enhanced PSO with the P&O, a more efficient hybrid MPPT of
Photovoltaics is created. The Search Skip Judge attempts to find an optimal solution by minimizing
the distance here between P-V curves (SSJ).

7
 3. Literature Review:

This incorporates a mechanism. As a result, the particle's unnecessary motion is minimized,

which expedites the convergence process. Known MPPT algorithms, such as customized InC,
hybrid PSO, as well as modified CS, are used to gauge the efficacy of the SSJ's initial version of
their proposed algorithm. Searching multiple required computation cycles, correctly determining the
global maximal power point, tracking time, and indeed the scanned interval are used to evaluate the
efficacy of these methods. The findings show that the proposed method significantly improves both
the precision and speed with which the true global total power can be determined. Not only that, but
it holds true regardless of the PS conditional complexity [10].

An examination of PS circumstances for PV energy technologies is detailed in [11]. When the

number of PV modules in a string increase in a PV electricity system operating under light load
conditions, the 0.8 Voc framework is used to improve efficiency. In the particular circumstance of
PS, the peaks are located at 0.8 Voc multiples. The authors [11] employed a novel model based on
a set of equations to determine the optimal placement of peaks in a PV energy system under partial
shading conditions. A model is used to solve the issue of intermittency in PV systems over the long
term. In contrast, the overarching PV system speed is slowed by this model.

Neural networks and fuzzy logic control are just two of the AI-based methods proposed by the
authors of [7]. The input combination in an ANN is interpreted as the environmental context for PV
parameters, and the expected output can be either a reference voltage or a duty cycle. A multi-level
neuro-fuzzy approach has been developed as an alternative to the standard neural network (NN)
method for MPPT. Under a wide variety of operational conditions, the prediction methodology has
been shown to be more efficient. [13] However, due to variations in PV array characteristics, the
system requires periodic training in order to maintain a high degree of accuracy [14].

The main advantages of Fuzzy Micro Controller (FLC) are its tolerance for a noisy input signal
as well as its convergence in the face of unexpected environmental change [15,16]. This method,
however, relied on a fuzzy rule - based table to identify the mistake. One major limitation of this
approach is its reliance on preexisting knowledge of something like the original signal [7]. The
Takagi-Sugeno model was applied in the literature to further enhance the system's stability and
performance [17,18]. F

8
 Researchers have developed a new MPP Flower Pollination Algorithm (FPA) that generates
five pollens or duty cycles with an FPA between 0 and 1, then calculates the average power of all
the individual pollens. The most potent pollen is recommended as the main power. To determine
whether a given pollen sample will go through global or local plant reproduction, a mobile address
between 0 and 1 is generated at random, with a switching probability "P" of 0.8 to achieve the
desired effect. It was determined after only 25 iterations [22] that the strongest power provides the
best global value. High computation duration, difficulty tuning parameters, complex system, and
procedural consequences are the primary drawbacks of FPA [23].

The MPP of a PV system with a modified form of FPA is described. Despite the system's
increased efficiency, FPA's structure has become more convoluted [24]. Using a technique for
tracking the MPP that relies on a random solution, Anderson et al. [25] analyzed the MPP in terms
of global optimization problems. Anderson et al. generate the solution haphazardly and seek out
suitable solutions in the interval 0-1 within the defined bounds. These numbers are fed into the DC-
DC converter to get a readout of the power at each input. Each iteration further reorganizes these
values again until MPP is reached.

In [26], a refined version of the Firefly algorithm (FA) is described that can be used to compute
MPP. Firefly randomness can be reduced by manipulating the parameters, which are introduced by
tinkering with the level of illumination and the temperature. By comparing their findings to those of
traditional FA, these same publishers of [26] were able to draw the conclusion that their method was
more effective and had a faster rate of tracking. G

9
3.1 Description of the Partially Shaded PV Energy System
3.1.1 Description of the Power and Control Circuits
The ideal photovoltaic cell is connected in parallel with a diode so that whenever light strikes
it acts like a current source. The characteristics of a photovoltaic module can be learned from a
theoretical model of something like a single diode. Parallel resistance (denoted by the symbol Rp)
and series resistance (denoted by Rs) make up a single diode model in practice. Figure 1 is a
schematic depiction of a PV model with a single diode.

"Figure 1. Single diode PV model"

Losses are shown by Rp and Rs in the single diode systematic procedure of a PV cell. Here are
the numerical methods of this model, which are based on Kirchhoff's current law (KCL):

I = Ipv − Id − Ish (1)

where (2) and (3) symbolize the values of The a as well as Ish below:

V + IRs
I−
od = I [exp( )]
(2)
nVT

Ish = (V + IRs)
R
sh
H

10
 where Io stands for the saturation current in the reverse direction. What VT is
worth in terms of is (3)

VT = Ns AKT
q

(4)
Ns is the absorption coefficient of the diode, K seems to be the Boltzmann constant, but
also T is the thermal efficiency of the photovoltaic (pv) cells.
As a result, we can see the full governing equation for a PV cell as well
as its output dc (I) as

+ IRs (V
I = Ipv − Io [exp(V ) − 1] −
(5)
+ IRs )

nVT Rsh H

11
 In photovoltaic energy systems, the DC-DC converter is an essential component of the final
product. A PV system's output power can be adjusted with the help of various converters, including
buck, boost, but instead buck-boost converters. Depending on the needs of the system, they either
lower or raise the voltage value. The dc - dc converter really does have the advantage of voltage
regulation, power, but also efficiency when compared to other Power converters. The output
voltages of PV energy systems should be regulated by changing the duty cycle of something like a
DC-DC boost converter. This requires coordinated operation of the terminal voltage and the load
being used. The DC to DC stimulate converter's basic schematic is shown in Figure 2.

"Figure 2. DC to DC Boost-Converter in PV system"

The boost converter's Vre f r has always been modulated to alter the PV voltage, which in
turn modifies the sinusoidal waveform (Vre f r) at the PV system's GM. The output frequency is
measured between 0 and 1, inclusive. Using a microcontroller with zero oscillation to implement
the control algorithm allows for maximum energy harvesting from the PV system. Below in (6)-(8)
are the mathematical representations of the boost converter's power dissipation, duty cycle, input
characteristic impedance, output power, as well as inductance (10).
∆Iripple Tswitching
Cinput
=
8∆Vripple,pv
fswitching

Ton
I
DC =
12
 (6)
(7)

I0
Coutput =
(8)
fswitching8∆Vrip
ple,load
(9)
Vin
Vout =
1 − DC (10)

L= DCIo

2 fswitching
∆ILoad

To calculate the duty cycle of something like a DC-DC boost converter, Vin is the voltage
applied, fswitching is the operating frequency, Iripple is the ripple current, but instead DC seems to
be an elaborate expression. Conventional methods in PV systems may be limited to LM and fail to
achieve GM, preventing the system from reaping its full potential in terms of energy production. As
a result, this study employs bio-inspired techniques to scour the MPP.
I

13
3.1.2 Description of UI, PS, and CPS under Study
When the irradiance on all PV panels is the same, we say that the panels are subject to
uniform irradiance. It detects changes in irradiance in a partially shaded environment. Because of
the PS, a string of interconnected PV modules may draw too little current and produce subpar
results. A string is a series connection of multiple PV modules that increases voltage. One inverter
is used for multiple strings.

"Figure 3. (a) Uniform Irradiance Case (b) Partial Shading Case"

When many PV modules are connected in series to produce an output, a condition known
as CPS exists [10]. The P-V but instead I-V characteristics exhibit multiple, nearby peaks. A cluster
is a group of three or more peaks in the power or voltage curves. Fig. 4 seems to be a block
diagram of the CPS.
J

14
"Figure 4. Complex Partial Shading Case"

In every group of CPS cases, there is only ever one GM. There is only one peak, and it is
called the cluster head peak. However, when looking at a PV curve from a PV energy system, the
number of peaks is always equal to the amount of partial shading conditions PV modules in a string.
If shading is dispersed evenly across PV energy systems, a distinctive P-V characteristics curve will
be displayed. And this curve has a number of closely spaced peaks. Locating GM is a major issue
in CPS.

3.1Previous MPPT Techniques for PV Energy System

3.1.1 "Incremental Conductance (InC) Technique"
The InC approach first determines the slope of conduction, after which the converter's duty
cycle is adjusted. The slope values can be calculated with this algorithm. The converter duty cycle
increases only for positive gradient values and decreases for all others. If the MPP is not found to
be on the curve, the procedure will be repeated [37]. When the generic version of Authority by
voltage is zero, that's when you know your PV array is operating at maximum power point
efficiency. The following are the frequently used InC fundamental equations:

devP
=0 (11)
devV

One possible notation for Equation (11) is as follows:

devP dev(I devI

V)
devV = I +V =0 (12)
devV devV
=
1
devI
<− at right o f MPP (13)
devV V
K devI 1
15
 >− at le f t o f MPP (14)
devV V
devI 1
=− at MPP (15)
devV V
Therefore, to get rid of something like the steady-state error, the authors of [38]
recommended an allowed by law error of or thereabouts = 0.002. The overall system's convergence
speed slows down and its cost goes up because of complexity [39].

3.1.2 Dragon Fly Optimization (DFO) Technique

Also, DFO is a biologically motivated algorithm. There are approximately 3,000 unique
species of dragonfly found throughout the world. The two stages of a dragonfly's lifecycle are
adulthood and the development into a nymph. They spend most of their time as a nymph
throughout their lives. After undergoing metamorphosis, they reach maturity. The dragonfly
looks at the little guys who eat big bugs. The swarming behavior they exhibit is extremely
unusual.

Like the two-stage meta-heuristics like exploitation and exploration, we find two behaviors.
Figure 5 displays the earliest known DFO pattern amongst humans.

"Figure 5. The primitive pattern of DFO Algorithm"

K
16
 Separation, placement, cohesion, food, but also enemy [40] are five behavior’s that can be
used to mathematically represent DF's movement within the swarm.
1. Separation. Any given DF in a deterministic swarm will not bump
into any other flies. Si is the distance between the ith people and
the rest of the population, and it can be calculated as
Si =
—∑N
(k=1) (P − Pk).
P is the current position of DF, and Pk is the kth position.

neighboring DF, and N has been the total population size.

2. Alignment. The other DF's speed is consistent with those of its
neighborhood residents, proving that they are not moving at a
different speed. The Chinese character for "alignment" (Ai) denotes.
∑N Vk
computed as Ai =N(k=1). Vk
the speed of the k nearest neighbors.
Cohesion refers to the tendency for people to gravitate toward the
center of a cluster of N Pk neighborhoods. Cohesion, denoted by Ci,
is defined as Ci = (k=1) P. Individuals in the DF are denoted by P, the
current position, and Pk, the kth position, of neighboring DF.
3. Food.
4. Everyone is making their way toward the food source as a matter of
survival. Food Fi is shown to be attracted to location x. Food (Fi) =
(PFood) + (P). The current person's position, denoted by P, and the
location of the currently consumed meal, denoted by PFood.
5. Enemy. Ei = PEnemy + P describes the collective behavior of those
fleeing from the aggressor. PEnemy represents the location of the
current foe, while P describes the position of something like the
protagonist. Individual DF positions are influenced by these five
characteristics. An individual's final upgraded DF position is found by
using the formula Pi = Pi + Pi. In the case of (16) below, where its
value of Pi is specified:

K ∆Pi = W∆Pi + (sSi + aAi + cCi + f Fi + eEi) (16)

17
4. Project (Proposed Hybrid (InC-DFO) Technique for PV Energy Systems):

4.1 Design:
Within a meta-heuristic framework, the researchers who wrote this paper combine the strengths
of InC with those of the variable step size but also DFO algorithm to create a new powerful
technique they call (InC-DFO).
The proposed InC-DFO method combines the benefits of these two algorithms into a single
computational strategy. These two methods are carried out in tandem to maximise the efficiency of
the optimization exercise. The DFO method's starting point is determined by the output of
something like the InC method, whose step size can vary. In contrast, the InC-DFO technique can
be used with PV energy systems in any climate, including UI, partial, as well as complex partial. In
the first stage of the InC-DFO algorithm, the InC methodology with variable step magnitude is used
due to its high performance under UI attributes and its low convergence as well as settling time, but
it gets stuck on LM, causing power loss.

dev × P
dk = d (k − 1) ± N .∗ (17)
d.ev
× V − dev × I

where d (k 1) and dk represent

— the duty cycle of the boost converter at time instant− k 1 and
× ×
time instant k, respectively. The×step size can be changed by changing the scaling factor, N. In
contrast, dev P, dev V, and dev I represent the PV array's actual power, voltage, and current
outputs. The correct form is as follows:
devP(k) = P(k) − P(k − 1) (18)
devV(k) = V(k) − V(k − 1) (19)

devI(k) = I(k) − I(k − 1) (20)

The modulation index at which the dc-to-dc increase converter operates specifies the
placement of the DFs during global MPP investigating of PV systems. This problem has only one
facet. The DA is used to increase the power output of the PV array, which is the objective function
to be optimized. The DF algorithm is applied in the 2nd stage of global MPP tracking and is
implemented as follows.
• Step 1: In terms of duty cycle, this same search space for the DFs
is contained within the range of dmin to dmax. In this article, we've
set dmin and dmax at 2% but rather 98%, respectively. The value
of Pi is set as the starting point for the population.
L • Step 2: The PV power of the DFs is determined for each position

18
 (duty ratio). This shifts both the location of food sources and
enemy forces.
• Step 3: Modified the ‘s’, ‘a’, ‘c’, ‘f’, and ‘e’ values.
• Step 4: The Equations Si (k=1)
= − ∑N (P − Pk) and Ei = PEnemy + P
are employed to determine an individual’s personal S, A, C, F, but also
E.
• Step 5: Adjustments have been made to the neighborhood's
perimeter.
• Step 6: The measurement Pi = Pi + Pi is used to alter the DF's
starting positions. If this equation yields a location beyond the dmin as
well as dmax bounds, then the flies will be placed beyond those limits.
• Step 7: If the algorithm's end condition is met, the process should
L end.ed.

19
Algorithm 1 Pseudocode of DF algorithm [40]

Make an n-by-Pi DF population I = 1, 2, n). Start off with Pi-step vectors I = 1, 2,..., n).
and the result is a disappointing situation.
Find out how the DF is doing overall. The enemies but instead food
supplies should be improved.
It is recommended to revise the current values of 'w,"s,' 'a,' 'c,' 'f,' and 'e'.
Determine the values of 'S,' 'A,' 'C,' 'F,' and 'E' by solving their corresponding
equations. Adjust the perimeter of the area.
if there is at least just one DF adjacent to A DF.
Change the velocity vector using Eq. (16). Calculate a new position vector
using the formula Pi = Pi + Pi.
else
The position vectors have been refreshed.
end if
Check the boundaries of the variables and modify the new positions accordingly.
M
end while

The proposed DF method for hybridization involves several variables that need to be adjusted.
Separation ('s'), alignment ('a'), cohesion ('c'), food ('f'), and enemy ('e') parameters can be used to
generate both exploitative and exploited behaviors during optimization. Some of these variables are
held constant while others shift with the various case studies. Neighboring DFs are crucially
important, so we assume that each fake DF has a neighborhood of a certain radius. Coordination of
flight while still allowing for sufficient separation and coherence is typical of a DF in a dynamic
swarm. Therefore, alignments are typically very weak, but cohesiveness is high enough to attack.

The duty cycle, which can take on values between 0 as well as 1, stands in for the search space in
our method. There is a connection between the output pulse width and indeed the DF's physical
location. Within the scope of the search, the sequence is obtained at random. Again, until the
stopping condition is met or the optimal solution is reached, the method phases are repeated, with
the exception of initiation. At GM, the fitness function is the most important source of PV power.

20
 From InC to DFO, a Partial Shading Detection Technique
Recognizing the I-V concave relationship of a silicon transistor PV cell is crucial to grasping the PS
detection technique of the InC-DFO method. Figure 6 depicts the full set of P-V curves for the PV
system, including UI, PS1, as well as PS2 conditions.

"Figure 6. Complete P-V curves of PV system"

An analogous circuit model is being investigated for this purpose. We derive Ipv's value from (5).
G
Ipv = (Ipv,STC + K∆T) (21)
GSTC

where K denotes the Boltzmann constant, T denotes the photovoltaic cell temperature, and G
denotes the irradiance. For a rewritten version of Ipv where T is held constant as well as G is
supposed to change, one could use the following form of the expression:

Ipv G
= (22)
Ipv,STC GSTC

Ipv,STC refers to the solar energy current at the industry-standard 1000 W/m2 and 25 °C
test conditions. Equation (22) can be derived from [41] if we assume that the region around
0.9Isc contains the region of maximum current Impp.

N Isc Impp G

21
 ≈ ≈ (23)
Isc,STC Impp,STC GSTC
Impp
G≈ STC G STC ≈
Isc
G (24)
Impp,STC Isc,ST
N

22
 Table 1. "Electric characteristics of Sun Power SPR-315E-WHT-D"

Description Electric Characteristics of

Sun Power SPR-315E-WHT-D

Maximum Power (Pmax) 315.072 W

Voltage at MPP (Vmax) 54.7 V

Current at MPP (Imax) 5.76 A

Short circuit current (Isc) 6.14 A

Open circuit voltage (Voc) 64.6 V

Temperature coefficient of (Voc) −0.176.6 mV/K

Temperature coefficient of (Isc) 3.5 mA/K

Temperature coefficient of power −0.38%/K

Peak efficiency 19.3%

O
23
"Figure 7. I-V and P-V curves of the PV system"
If we observe the ratio of Isc when G decreases from 1000 W/m2 to 700 W/m2, we find that
Isc 4.202
= = 0.684 ≈ 0.7 (25)
Isc,STC 6.14

In a similar vein, when the MPP current ratio changes and G decreases from 1000 W/m2 to
700 W/m2, the same conclusions could be drawn.

Impp 4.046
= = 0.689 ≈ 0.7 (26)
Impp,STC 5.867

From the (24), we can see that the irradiance values measured at Impp and Isc are very
close to one another. Impp is hidden close to the 0.8 Vocarray open-circuit voltage in [41], while Isc
is located close to the 0.8 Vocmodule open-circuit voltage. Therefore,
− we can express Impp as an
I0.8Voc array and Isc as an I0.8Voc. Therefore, (24) can also be expressed as:

I I
G ≈ 0.8Voc−array GSTC ≈ 0.8Voc GSTC (27)

O Impp,STC Isc,STC

24
 When a sample is taken at 0.8Voc array as well as 0.8Voc, MPPT logs the corresponding
magnetic fields as 0.8Voc and 0.8Voc. G values are determined at the 0.8Voc array, with G-1 and
G-2 being defined as the 0.8Voc values.

— I0.8Voc GSTC
G 1=
Isc,ST (28)
C
6.001
= × 1000 = 977
6.14

I
G 2—= 0.8Voc−arraySTC
G
Impp,STC (29)
5.7
8
= × 1000 = 1003
5.76

It can be seen from equations (28) but also (29) that G-1 as well as G-2 are very close to the true
value of G, which is around 1000 W/m2. The value of Impp is assumed to be 0.9 of Isc in (27)
though. Due to the potential for producing a mismatch value, this evaluation is not necessarily
correct for all moral standards of G. When comparing G-1 and G-2 at 1000 W/m2, the mismatch
value is 23 W/m2, as calculated above. At 300 W/m2, the G mismatch importance within G-1 and
G-2 would be reduced nearly to zero, while at 700 W/m2 it is lowered to 12 W/m2.

25
 |G1 − G2| > 40PS
Conditi
on (30)
|G1 − G2| < 40UI
Conditi
on

In the second phase, the bioinspired DFO method is utilized to search the GM and solve the
power loss issue. It is imperative that the PV energy source converge to 80% of Isc in the first
stage, and an InC technique with predictive control size is used to achieve this. The DFO algorithm
is triggered when the InC algorithm reaches the Isc, and it settles this same PV array to the precise
MPP. So, not only does InC-DFO search the MPP under PS and CPS conditions, but it also
eliminates the drawbacks of both methods. Figure 8 presents a flowchart describing the proposed
InC-DFO algorithm.
P

26
"Figure 8. Flowchart of Proposed Hybrid (InC-DFO) Technique"

4.2 Simulation Model

FIGURE 9: Common PV system design showing MPPT methods in action. The framework
for the whole setup is the PV model described in [43]. A PV array is made up of four modules
connected in series. This PV module has the following specifications: Pmpp = 200 W, Vmpp =
26 V, Voc = 32.9 V, Imp = 7.6 A, Isc = 8.21. A PV array experiencing varying degrees of partial
shading is used to evaluate the effectiveness of the newly developed Hybrid (InC-DFO) based
MPPT. A boost converter links the photovoltaic array to the battery load. According to [44], the
DC-to-DC converter can be optimized with the following values: Cin = 10 uF, L = 200 uH, Cout
= 47 uF, but instead Fsw = 20 kHz. The MPPT procedures have been implemented in
embedded blocks, as shown in Figure 9. This DC-to-DC boost converter duty cycle-outputting
block takes Ipv as well as Vpv as inputs.

27
28
 Figure 9. Simulation Model.

R
4.2 Results and Analysis:
All-weather scenarios (UI, PS, and CPS) are covered in this section's four cases. One case is
selected for UI, couple for PS, one for CPS out of a total of four. Four outlying cases with varying
irradiance values are selected based on the results and trials of the algorithms inspired by nature.
The MATLAB/Simulink software has been used to create a model of PV energy system. Table 1
shows the power output of a solitary PV module used in a model of a PV energy system.

29
30
 The developed PV model is put through its paces using a battery of previously developed
algorithms in addition to the experimental MPPT Hybrid (InC-DFO) method. All MPPT methods for
PV distributed generators are put through their paces across four different use cases, with the UI
condition serving as an example in the first. The second and third examples detail the PS condition.
It is in this fourth scenario that the CPS condition becomes explicit. The data is analyzed based on
settling time, current, voltage transients, power, efficiency, but instead oscillation. A breakdown of
all four scenarios and their outcomes is provided below.

4.3.1 Case 1. Uniform Irradiance Condition

In this case, the irradiance received by all of the PV system's modules was consistent over
time but varied. Table 2 summarizes the irradiance pattern. The P-V and I-V UI curves for PV
energy systems are shown in Figure 7. Table 2 provides more information about the uniform
irradiance pattern that is used.
Among the techniques tested, the proposed hybrid (InC-DFO) method was found to have
the highest power under uniform irradiance conditions, reaching 863.4 W at 700(KW) irradiance
levels. When compared to other methods, FA uses the least amount of power, just 852 W. Hybrid >
DFO > InC > CS > ACO > PSO > FA is the order of power as it is monitored. The CS algorithm
exhibits greater oscillation and has the longest time to be set at 1.94 s. FA measured the lowest
power consumption, at only 852 W of any of the methods. By dampening vibration and increasing
stability, the proposed hybrid method improves PV system efficiency by a whopping 99.93 percent,
saving a ton of power in the process. Compared to other methods, CS's power-draining oscillations
are the highest. R

31
"Table 2. P-V and I-V characteristic curves at different irradiance values"

M2

Case Irradiance Si( KW ) PowerMax

PV-1:0.70 PV-2:0.70 PV-3:0.70

Case 1 UI PV-4:0.70 864 W

PV-1:1.00 PV-2:0.80 PV-3:0.10

Case 2 PS1 PV-4:0.20 512 W

PV-1:0.90 PV-2:1.00 PV-3:0.80

Case 3 PS2 PV-4:0.50 781 W

PV-1:1.10 PV-2:0.25 PV-3:0.84

PV-5:1.00
Case 4 CPS PV-4:0.40 PV-6:0.50 1250 W
PV- PV-8:0.40 PV-9:0.80
7:0.90
PV-
10:0.20

Reducing the step size value, which can lengthen the tracking time, can help mitigate the
oscillation. To strike a good balance between efficiency as well as tracking time, this same DC to
DC boost converter's step size should be between 2 and 5 percent of the power from the PV arrays.
Large fluctuations are brought about by levy flight in CS. It was found that the hybrid method was
the most effective, had no oscillations, tracked quickly, had the shortest convergence time at 0.50
seconds, and had the longest settling time at 0.59 seconds.

4.3.2 Case 2. Partial Shading Condition

Table 2 shows the irradiance pattern under partial shading conditions discussed in the
context of Case 2. Whenever PV arrays don't collect the same amount of irradiance from the sun,
they produce less than optimal results. Less-efficient diodes caused by PS require bypass diodes.
The search for the peaks can be made more difficult by the fact that both the I-V and P-V
characteristics curves feature multiple peaks. Furthermore, a cluster of peaks can be described as a
collection of LMs and a single GM.
S
32
"Figure 10. P-V and I-V Characteristics Curves for Partial Shading Cases 2 and 3" S

33
 Traditional InC methods exhibited the best power tracking. PSO, DFO, InC, Hybrid, CS, FA,
as well as ACO all take about 1.80 seconds to reach stability; InC takes 0.75 seconds; Hybrid takes
1.78 seconds; CS takes 1.30 seconds; and ACO takes 0.93 seconds. Figure 11 shows a
comparison of power, current, and voltage across all methods for Cases 1 and 2.

"Figure 11. Power, Current and Voltage comparison of all techniques for Case 1 and Case 2"

34
4.3.2 Case 3. Partial Shading Condition

A third PS condition is illustrated in Table 2 by the

× irradiance pattern of Case 3. In this
instance, PS is used to monitor GM's worth. So far, we have calculated that 781 W is the value of
GM. A 4 1 PV array was being used if there is partial shading. DFO, PSO, CS, FA ACO, and InC
are all compared to the proposed hybrid method to see how it stacks up. Even though PSO has the
greatest frequency deviation and ACO has none, both methods are stuck on LM and cannot locate
the GM. When the PV system vibrates, energy is wasted.

Various optimization methods have different average convergence times; these are 1.10 s
for PSO, 0.53 s for DFO, 0.22 s for InC, 0.20 s for CS, 1.41 s for FA, and 0.31 s for ACO. InC is
shown to have the fastest convergence time among all the methods. ACO, FA, CS, hybrid, InC,
DFO, and PSO each have a settlement time of 0.35 s, 1.42 s, 1.93 s, 0.89 s, 0.22 s, 0.73 s, as well
as 1.65 s, including both. The proposed method has the shortest settling time, which reduces power
losses and boosts PV energy systems' effectiveness. When it comes to potential power outages
caused by settling time, CS ranks highest.
T

35
4.3.2 Case 4. Complex Partial Shading Condition
Table 2 depicts the irradiance pattern for Case 4 of something like the CPS situation.
In Figure 12, we can see the P-V and I-V characteristics curves of CPS events.

"Figure 12. P-V and I-V Characteristics Curves for Complex Partial Shading Cases"

U
The hybrid algorithm's 99.98% success rate in CPS was the greatest of any algorithm.

36
 maximum, but the GM eludes my search. Accordingly, the DFO with the most powerful tracking
capability is used in the second phase. The suggested hybrid method improves PV system
efficiency since it oscillates at zero and converges at the fastest possible rate. Figure 13 shows a
comparison of power, current, as well as voltage across all methods for Cases 3 and 4.

"Figure 13. Power, Current, and Voltage comparison between all techniques for Case 3 and Case
4."

Four distinct cases involving Photovoltaic array have been taken for any and all-weather
circumstances to validate the efficiency of the recommended (InC-DFO) itself and comparison using
six other MPPT strategies. One case represents the UI scenario, two represent the PSC scenario,
and one represents the CPS scenario. Table 3 displays the results of the numerical analysis
performed on these four scenarios to determine the PV system's performance valuation.

37
 W

Table 3. "Comparative Analysis between Proposed and Previous MPPT Techniques for all
Weather Conditions"

Case Techni Converge Settlin Maximu Power Efficien GM

que nce g m cy
Time (s) Time Power Track (%) De-
(s) ed
(Watt) (Watt) tected
PSO 0.80 1.10 864 862 99.77 Yes
DFO 0.53 0.61 864 863.3 99.92 Yes
InC 0.52 0.60 864 863.1 99.90 Yes
1.UI Hybrid 0.50 0.59 864 863.4 99.93 Yes
C
CS 1.90 1.94 864 863 99.88 Yes
FA 0.60 0.67 864 852 98.61 Yes
ACO 0.53 0.62 864 862.2 99.76 Yes
PSO 1.60 1.80 512 498 97.27 Yes
DFO 0.59 0.75 512 509 99.41 Yes
InC 0.48 0.52 512 325 63.48 No
2.PS Hybrid 0.48 0.62 512 511.4 99.88 Yes
1
CS 1.60 1.78 512 506 98.83 Yes
FA 1.00 1.30 512 508 99.22 Yes
ACO 0.65 0.93 512 507 99.02 No
PSO 1.10 1.65 781 779 99.74 Yes
DFO 0.53 0.73 781 780 99.87 Yes
InC 0.22 0.22 781 712 91.17 No
3.PS Hybrid 0.20 0.89 781 780.4 99.92 Yes
2
CS 1.41 1.93 781 770.2 98.59 Yes
FA 1.01 1.42 781 769 98.46 Yes
ACO 0.31 0.35 781 768 98.33 No
PSO 1.31 1.91 1250 1246 99.68 Yes

38
 DFO 0.69 0.83 1250 1248 99.84 Yes
InC 0.39 0.45 1250 1094 87.52 No
4.CP Hybrid 0.41 0.75 1250 1249.8 99.98 Yes
S
CS 0.73 1.82 1250 1243 99.44 No
FA 0.52 0.80 1250 1157 92.56 No
ACO 0.59 1.40 1250 1247.1 99.76 Yes

The methods used in soft computing have been found to be reliable, redundant, and
extremely flexible. While the conventional InC conductance method has the quickest convergence
time in clear skies, it loses the GM in a couple of partial-shade scenarios. The algorithm's efficiency
stems from its gradient-based control but also its simplicity, as well as its low-cost control. Only
under the UI scenario, when all PV modules have the same irradiance values and GM can be
found, are InC approaches practical. However, when compared to PS, PSO, Hybrid, CS, FA, as
well as ACO, this traditional method has the worst efficiency.

The DFO methodology is included in the comparative analysis of the offered methods because
it contains useful, distinguishing features such as the ability to avoid and cohere with an enemy's
attacks and the ability to engage with closely- spaced swarms. These DFO features work to
dampen vibrations and display the best possible PV energy system performance. When compared
to other methods, the hybrid approach is clearly the most effective.

There is no chance of oscillation in their output under any circumstances. Both the transient
and steady-state responses can be handled by the hybrid method.
W

39
 When compared to the results presented in the study [45], the results demonstrate improved
performance. Power tracked by the proposed hybrid and every other DFO, CS, FA, InC, as well as
ACO techniques is 511.4 W, 509 W, 506 W, 498 W, 325 W, as well as 507 W, respectively, and
settling times for these methods range from 1.80 seconds to 0.75 seconds, 0.52 seconds to 0.62
seconds, 1.78 seconds to 1.30 seconds, and 0.93 seconds for the PS1 condition. PSO, DFO, InC,
InC-DFO, CS, FA, as well as ACO all have different PS2 condition convergence times: 1.10s, 0.53s,
0.22s, 0.20s, 1.41s, 1.01s, and 0.31s. Of all the methods, it is the hybrid algorithm that provides the
most strength. In Table 3, you can see that its convergence and settling times are the shortest. As a
result of their two-stage running phenomena, the suggested algorithm effectively follows the GM
and provides highest efficiency of 99.93%, 99.88%, 99.92%, but instead 99.98% under all operation
conditions. W
4.3.2 Case 5. Field Atmospheric Data of Beijing
The practicality of the PV system is demonstrated by several indicators. Open areas are
frequent locations for solar PV systems, making them vulnerable to inclement weather. In Case 5,
we conduct a field investigation in Beijing, the capital of China. Evaluation of the recommended InC-
DFO MPP trying to search method using field atmospheric data. The Chinese Meteorological
Administration provides seasonal temperature and radiation data [46]. All the day's incoming solar
irradiance and temperature are calculated throughout two seasons. This PV installation features
four 315.072-watt arrays, for a total system output of 1260.288-watts. Data on the solar irradiation
and temperature of something like the PV system are included. Fluctuations in the resolution of the
24-hr is 10 minutes. Effectiveness of the Mppt control InC-DFO algorithm is measured by the
maximum power produced during a 24-hour period.

Outcomes for Case 5 are presented in Table 4. The peak production power in Beijing is 897.2
W in the springtime, between 10 a.m. and 3 p.m., making the summer the most productive season.
The average peak power in the fall is 811 W, and it occurs at the earliest hour of the season.
Seasons of spring and winter seem to be the least productive seasons for acquiring power. Beijing
may be found at coordinates 39.9042 N and 116.4074 E. The InC-DFO algorithm performed well in
all conditions, the results showed. Figures 14 and 15 provide a detailed analysis of the power
acquired by the InC-DFO algorithm in Beijing during the course of 24 hours in the summer and
spring using ACO, FA, DFO, InC, PSO, as well as CS, respectively. The outcomes prove the
superiority and efficiency of the proposed InC-DFO algorithm.
X

40
Figure 14. "Power comparing of Spring in Beijing of InC–DFO algorithm with CS, FA, ACO,
PSO, InC, and DFO"

Figure 15. " Power comparing of Summer in Beijing of InC–DFO algorithm with CS, FA, ACO,
PSO, InC, and DFO"

Table 4. "Results summary of case 5 field atmospheric data in Beijing City"

Name of Seaso Calculate In CS FA AC PS InC DFO

City n C- O O
DF
O
Beijing Sum Powermax 112 109 110 110 11 109 111
mer (W) 2 8 7 9 02 3 3
Spri Powermax 110 108 108 109 10 108 109
Y ng (W) 5 5 9 5 87 0 8

41
5 Conclusions
To extract the highest possible energy from solar PV systems, this study details a composite
(InC-DFO) MPPT method. Six alternative methods, including PSO, DFO, FA, ACO, InC using
adaptive threshold size, and CS MPPT, are compared to and contrasted with the presented InC-
DFO method. One case represents the UI condition, two represent the PS condition, and one
represents the CPS condition in this comparison across four different cases. Analytical analysis and
statistical data from Table 3 demonstrate that the suggested hybrid technique exceeds the state-of-
the-art in terms of number of iterations, settling time, power consumption, efficiency, and resilience.
Even when accounting for CPS, its 0.75-second settling time for locating the GM is the quickest
available. The InC-DFO method provides the best performance across a wide range of climates and
climatic situations.

This proposed technique saved the power losses because of the fast-searching ability of a
GM and reduced the settling time that may enhance the performance of the PV energy systems in
a transient state. Moreover, ripples and overshoot is eliminated because of zero oscillation in the
transient state. It is thus concluded that although the conventional techniques are suitable for
uniform irradiance, they do not find GM in the case of partial shading. Similarly, they require less
time for tracking, but they are stuck on LM. On the other side, bio-inspired methods are complex
and take more time in finding the MPP. The proposed hybrid technique is the combination of InC
and DFO, and it is used in two stages. In the first stage, the conventional InC technique with
variable step size is adopted because of less convergence and settling time while in the second
stage, a bioinspired DFO technique is implemented for finding the GM quicker. The proposed
technique outperforms the competition in terms of MPP monitoring and integration at the global
MPP. The suggested MPPT technique efficiently achieves tracking time improvements of 10–29%
and spontaneous oscillation reductions of more than 91%. The hybrid MPPT technique provides
faster maximum power point (MPP) monitoring, lower computational pressure, and increased
performance.
Z

42
6 References
AA
1. Pourbehzadi, M.; Niknam, T.; Aghaei, J.; Mokryani, G.; Shafie-khah, M.; Catalão, J.P.
Optimal operation of hybrid AC/DC microgrids under uncertainty of renewable energy
resources: A comprehensive review. Int. J. Electr. Power Energy Syst. 2019, 109, 139–159.
[CrossRef]
2. Kumar, N.; Nema, S.; Nema, R.K.; Verma, D. A state-of-the-art review on conventional,
soft computing, and hybrid techniques for shading mitigation in photovoltaic applications.
Int. Trans. Electr. Energy Syst. 2020, 30, e12420. [CrossRef]
3. Pachauri, R.; Singh, R.; Gehlot, A.; Samakaria, R.; Choudhury, S. Experimental analysis to
extract maximum power from PV array reconfiguration under partial shading conditions. Eng.
Sci. Technol. Int. J. 2019, 22, 109–130. [CrossRef]
4. Saravanan, S.; Babu, N.R. Maximum power point tracking algorithms for photovoltaic
system—A review. Renew. Sustain. Energy Rev. 2016, 57, 192–204. [CrossRef]
5. Mao, M.; Cui, L.; Zhang, Q.; Guo, K.; Zhou, L.; Huang, H. Classification and
summarization of solar photovoltaic MPPT techniques: A review based on traditional and
intelligent control strategies. Energy Rep. 2020, 6, 1312–1327. [CrossRef]
6. Huang, Y.P.; Hsu, S.Y. A performance evaluation model of a high concentration
photovoltaic module with a fractional open circuit voltage-based maximum power point
tracking algorithm. Comput. Electr. Eng. 2016, 51, 331–342. [CrossRef]
7. Javed, M.Y.; Gulzar, M.M.; Rizvi, S.T.H.; Arif, A. A hybrid technique to harvest maximum
power from PV systems under partial shading conditions. In Proceedings of the 2016
International Conference on Emerging Technologies (ICET), Islamabad, Pakistan, 18–19
October 2016; pp. 1–5.
8. Bhukya, M.N.; Kota, V.R. A quick and effective MPPT scheme for solar power generation
during dynamic weather and partial shaded conditions. Eng. Sci. Technol. Int. J. 2019, 22,
869–884. [CrossRef]
9. Costanzo, L.; Vitelli, M. Resonant electromagnetic vibration harvesters applications:
Optimization of P&O MPPT technique parameters. In Proceedings of the 2018 Thirteenth
International Conference on Ecological Vehicles and Renewable Energies (EVER), Monte
Carlo, Monaco, 10–12 April 2018; pp. 1–8.
10. Kermadi, M.; Salam, Z.; Ahmed, J.; Berkouk, E.M. An effective hybrid maximum power
point tracker of photovoltaic arrays for complex partial shading conditions. IEEE Trans. Ind.
Electron. 2018, 66, 6990–7000. [CrossRef]

43
AB
11. Ahmed, J.; Salam, Z. An improved method to predict the position of maximum power point
during partial shading for PV arrays.
IEEE Trans. Ind. Inform. 2015, 11, 1378–1387. [CrossRef]
12. Chin, V.J.; Salam, Z. A new three-point-based approach for the parameter extraction of
photovoltaic cells. Appl. Energy 2019,
237, 519–533. [CrossRef]
13. Kumar, M.S.; Manoharan, P.; Ramachandran, R. Modelling and simulation of ANFIS-
based MPPT for PV system with modified SEPIC converter. Int. J. Bus. Intell. Data Min.
2019, 15, 255–272. [CrossRef]
14. Kayri, I.; Gencoglu, M.T. Predicting power production from a photovoltaic panel through
artificial neural networks using atmospheric indicators. Neural Comput. Appl. 2019, 31,
3573–3586. [CrossRef]
15. Hafendi, M.A.; Abd Wahab, H. Optimization of Stand-Alone Polycrystalline Photovoltaic
Array System by Implementing Fuzzy Logic and Perturbation and Observation Maximum
Power Point Tracking Controller. Commun. Comput. Appl. Math. 2019, 1, 34–40.
16. Priyadarshi, N.; Sharma, A.K.; Bhoi, A.K.; Ahmad, S.; Azam, A.; Priyam, S.
MATLAB/Simulink based fault analysis of PV grid with intelligent fuzzy logic control
MPPT. Int. J. Eng. Technol. 2018, 7, 198–204. [CrossRef]
17. Khabou, H.; Souissi, M.; Aitouche, A. MPPT implementation on boost converter by using
T–S fuzzy method. Math. Comput. Simul. 2020, 167, 119–134. [CrossRef]
18. Farajdadian, S.; Hosseini, S.H. Design of an optimal fuzzy controller to obtain maximum
power in solar power generation system.
Sol. Energy 2019, 182, 161–178. [CrossRef]
19. Hong, Y.Y.; Beltran, A.A., Jr.; Paglinawan, A.C. A robust design of maximum power point
tracking using Taguchi method for stand-alone PV system. Appl. Energy 2018, 211, 50–
63. [CrossRef]
20. Ram, J.P.; Babu, T.S.; Rajasekar, N. A comprehensive review on solar PV maximum
power point tracking techniques. Renew. Sustain. Energy Rev. 2017, 67, 826–847.
[CrossRef]
21. Shahdadi, A.; Khajeh, A.; Barakati, S.M. A new slip surface sliding mode controller to
implement MPPT method in photovoltaic system. In Proceedings of the 2018 9th Annual
Power Electronics, Drives Systems and Technologies Conference (PEDSTC), Tehran,
Iran, 13–15 February 2018; pp. 212–217.
22. Oliveira-Pinto, S.; Stokkermans, J. Assessment of the potential of different floating solar
44
 technologies–overview and analysis of different case studies. Energy Convers. Manag.
2020, 211, 112747. [CrossRef]
23. Ram, J.P.; Pillai, D.S.; Ghias, A.M.; Rajasekar, N. Performance enhancement of solar PV
systems applying P&O assisted Flower Pollination Algorithm (FPA). Sol. Energy 2020,
199, 214–229.
24. Pei, T.; Hao, X.; Gu, Q. A novel global maximum power point tracking strategy based on
modified flower pollination algorithm for photovoltaic systems under non-uniform irradiation
and temperature conditions. Energies 2018, 11, 2708. [CrossRef]
25. Kostenko, V.; Seleznev, L. Random Search Algorithm with Self-Learning for Neural Network
Training. Opt. Mem. Neural Netw.
2021, 30, 180–186. [CrossRef] AB

45
 7 Appendix

function duty = MPPT_algorithm(vpv,ipv,delta)

% Initialize algorithm parameters and define maximum a minimum
% values of duty cycle
duty_init=0.1;
duty_min=0;
duty_max=0.85;
% Define internal values for the voltage and power as persistant % variables
persistent Vold Pold duty_old;
% Initialize the internal values for the power and voltage on the % first pass
if isempty(Vold)
Vold = 0;
Pold = 0;
duty_old = duty_init;
end
% Calculate measured array power
P = vpv*ipv;
dV= vpv - Vold;
dP= P - Pold;
% Increase or decrease duty cycle based on if dP ~= 0 && vpv>30
if dP < 0
if dV < 0
duty = duty_old - delta;
else
duty = duty_old + delta;
end
else
if dV<0
duty = duty_old + delta;
else
duty = duty_old - delta;
end end
else
duty=duty_old;
end
% Check duty cycle values
if duty >= duty_max duty=duty_max;
elseif duty < duty_min duty=duty_min;
end
% Update internal values
duty_old=duty;
Vold=vpv;
Pold=P;

2
3