INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
Proposition of Ant Colony and Perturb and Observe MPPT  
Combination for Photovoltaic System  
Kharismi Burhanudin1*, Mohd Nazrien Zaraini2, Muhammad Faheem Mohd Ezani3, Muhamad Nabil  
Hidayat4,  
1,2,3Centre for Advanced Computing Technology (C-ACT), Fakulti Kecerdasan Buatan dan  
Keselamatan Siber (FAIX), Universiti Teknikal Malaysia Melaka (UTeM), 76100 Durian Tunggal,  
Melaka, Malaysia  
4College of Engineering, UiTM Shah Alam, Selangor, Malaysia  
*Corresponding Author  
DOI: https://dx.doi.org/10.47772/IJRISS.2025.910000725  
Received: 01 November 2025; Accepted: 08 November 2025; Published: 22 November 2025  
ABSTRACT  
This research presents a novel hybrid Maximum Power Point Tracking (MPPT) technique that combines Ant  
Colony Optimization (ACO) with the Perturb and Observe (P&O) method to enhance the efficiency and  
convergence speed of photovoltaic (PV) systems. The ACO MPPT, based on swarm intelligence, is utilized for  
its ability to conduct a global search for the maximum power point. In contrast, the P&O method provides  
steady-state tracking with low computational complexity. By integrating these two approaches, the research  
aims to leverage their respective strengths to achieve faster and more reliable convergence under varying  
environmental conditions. The study employs the NTR 5E3E monocrystalline PV module (173.5W) as the test  
subject, with the implementation carried out in a MATLAB Simulink environment. The experimental results  
demonstrate that the hybrid approach outperforms the standalone P&O and ACO MPPT methods in terms of  
convergence speed, accuracy, and stability, indicating promising potential for practical applications in PV  
systems.  
Keywords— Aco, Pso, Fpo, Pno, Mppt, Pv, Matlab Simulink  
INTRODUCTION  
A photovoltaic (PV) system converts sunlight into electrical energy and plays a crucial role in renewable  
energy generation [1]. The efficiency of a PV system largely depends on effectively extracting maximum  
power from the PV panels [2]. This is achieved through the integration of essential components such as PV  
panels, power converters, and maximum power point trackers (MPPT). The MPPT is vital for ensuring that the  
PV system operates at its optimal point, maximizing energy harvest under varying environmental conditions.  
This research explores a hybrid MPPT method that combines two distinct techniques: Ant Colony  
Optimization (ACO) and Perturb and Observe (P&O). Both methods aim to track the maximum power point  
(MPP) but differ significantly in their operational mechanisms. The P&O method is a simple and widely used  
technique that adjusts the operating point based on the slope of the PV power curve, offering ease of  
implementation [3], but it is known to suffer from oscillations around the MPP [4]. In contrast, the ACO  
algorithm, inspired by swarm intelligence, performs a global search across the solution space by mimicking the  
foraging behavior of ant colonies [5]. While it offers enhanced exploration capabilities, it also comes with  
increased computational complexity. Each method has its own advantages and limitations: P&O is fast but can  
oscillate near the MPP, whereas ACO is robust but requires greater computational resources. This study  
investigates the potential of combining both methods to develop an effective hybrid MPPT strategy [6]. The  
goal is to leverage the rapid convergence of P&O alongside the global search efficiency of ACO, while  
addressing the common trade-offs found in traditional MPPT techniques. The system setup utilizes a boost  
Page 8814  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
converter due to its efficiency and capability to manage high-voltage conversion requirements. Experimental  
evaluations are conducted using a practical 173.5W NTR 5E3E monocrystalline PV module, highlighting the  
real-world applicability of the proposed hybrid approach. This research systematically analyzes the  
performance of individual MPPT methods and their combination, providing insights into parameter tuning and  
the feasibility of implementing this hybrid technique in cost-effective PV systems. Ultimately, the study  
demonstrates the benefits of using soft computing methods, which are valued for their simplicity and low-cost  
implementation, in optimizing PV energy harvesting performance.  
METHODOLOGY  
This section provides a comprehensive overview of advanced techniques for maximum power point tracking  
(MPPT) in photovoltaic (PV) systems, organized into four distinct subsections. The first subsection discusses  
the impedance matching process and how to calculate the operating region to achieve maximum power  
transfer, which is essential for the efficient operation of PV systems [1], [2]. The second subsection details the  
implementation of the Ant Colony Optimization (ACO) algorithm for MPPT [7], demonstrating how bio-  
inspired computational methods can enhance energy extraction [5]. The third subsection explores the Perturb  
and Observe (P&O) method [3], a widely used and straightforward MPPT technique that modifies system  
variables to locate the maximum power point [8]. Finally, the fourth subsection presents a hybrid approach that  
combines the ACO and P&O algorithms, aiming to leverage the strengths of both methods for improved  
performance in MPPT for PV systems [9]. Together, these subsections provide an in-depth examination of  
both conventional and innovative MPPT strategies [10].  
A. Impedance Matching Process and Operating Region Calculation for Maximum Power Point  
Tracking  
The impedance matching process is a method used to determine the maximum power that a photovoltaic (PV)  
system can produce at a specific irradiance level. This method serves as a reference for analyzing PV power  
output. The impedance matching process is based on a boost converter circuit, where equivalent resistance,  
load resistance, and duty cycle are key factors in the impedance matching calculations [1]. Table I presents the  
NTR 5E3E PV module used in this research study [1].  
TABLE I The Ntr 5e3e Pv Module Parameter [11].  
Parameters  
Value  
44.4V  
5.4A  
Open circuit voltage,  
Short circuit current,  
Maximum Voltage,  
Maximum Current,  
Maximum Power,  
35.4V  
4.9A  
173.5W  
Each formula for calculating impedance matching in converters is different. The formula for impedance  
matching used in boost converters is listed below:  
(
/
) = (  
/
) = 1/(1 −  
)
(1)  
(2)  
=
(1 − ) 2  
1/2  
= 1 − (  
Where  
/
)
(3)  
is Equivalent Resistance,  
is Load Resistance and  
is Duty cycle [12]. The research study  
2
focusses on varying irradiance from 100-1000W/m which is sufficient to provide analysis for this research  
work. The identification of the operating region is necessary to determine the active region for the boost  
Page 8815  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
converter to function with MPPT [13]. Table II shows the equivalent resistance and duty cycle data for  
irradiance from 100-1000W/m 2 based on NT5E3E PV panel.  
TABLE II The Duty Cycle Data Based on Varying Irradiance Selection.  
Irradiance  
1000W/m 2  
900W/m 2  
800W/m 2  
700W/m 2  
600W/m 2  
500W/m 2  
400W/m 2  
300W/m 2  
200W/m 2  
100W/m 2  
V max  
35.4  
I max P max  
4.9 173.5  
R eq  
7.22  
8.03  
9.03  
10.39  
12.23  
14.87  
19  
D. cycle  
0.6532  
0.6342  
0.612  
35.4  
4.409 156.1  
3.911 138.4  
3.406 120.6  
2.896 102.5  
2.381 84.3  
1.863 65.95  
1.341 47.47  
0.879 28.64  
0.416 9.77  
35.4  
35.4  
0.584  
35.4  
0.544  
35.4  
0.5023  
0.437  
35.4  
35.4  
26.4  
37  
0.337  
32.57  
23.52  
0.215  
58.32  
0.014  
According to Table II, the load resistance must be greater than 58.32 ohms to efficiently track maximum power  
across various irradiance levels from the PV panel [12]. Therefore, a load resistance value of 60 ohms is used  
for the PV panel.  
Inclination angle formula  
2
Ѳ
=
−1[1/ (1 −  
)
]
(4)  
Parameters description:  
Ѳ
: Inclination Angle  
: Duty Cycle  
: Load Resistance  
The data presented illustrates the relationship between duty cycle and the inclination angle (Ѳ), ranging from a  
maximum of 90° to a minimum of approximately 0.98° [13]. As the duty cycle decreases from 1 to near zero,  
the inclination angle correspondingly declines from its maximum value of 90° to its minimum value of  
approximately 0.98°.  
TABLE III The Operating Region of Boost Converter for Mppt Process.  
Duty Cycle  
1
Inclination Angle, Ѳ  
90 (Ѳ  
7.89  
)
0.6532  
0.6342  
0.612  
7.099  
6.314  
5.50103  
4.583  
0.584  
0.544  
Page 8816  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
0.5023  
0.437  
0.337  
0.215  
0.014  
3.8503  
3.01  
2.172  
1.55  
0.9822 (Ѳ  
)
Accurately identifying this range is crucial for efficient system performance, especially in contexts such as  
maximizing power transfer or adapting to environmental conditions. The optimal operation likely occurs  
within this defined zone, where small variations in the duty cycle significantly influence the inclination angle,  
thereby affecting the system's energy conversion efficiency. This analysis highlights the importance of  
precisely determining the operating region to ensure the system operates at peak efficiency while maintaining  
stability across varying duty cycle conditions [3], [4].  
Fig. 1 The operating region of boost converter for MPPT process [14].  
Implementing Ant Colony Algorithm for MPPT  
Ant Colony MPPT is implement by making most out of ant’s behavior [7]. The concentration of the  
pheromone from the initial ants will affect the overall ant movement during the swarm process. Revised  
concentration of the pheromone, change in pheromone concentration and Pheromone concentration rate of the  
ants ill affect the whole maximum tracking process of the PV power. Listing below show the characteristic of  
the ACO algorithm formula [5]:  
The pheromone concentration formula  
Tij = ρTij t − 1 + ΔTij  
(5)  
Parameters description:  
: Revised concentration of pheromone  
: Change in pheromone concentration  
: Pheromone concentration rate (0-1)  
The initialization of pheromone concentration rate is crucial due to the impact on convergence process of the  
PV power. If the Pheromone concentration is wrongly initialized, it will cause false convergence and easily  
causes the convergence fall to local minimum. Fig. 2 shows the ACO MPPT technique that leads to MP  
identification process.  
Page 8817  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
Fig. 2 ACO MPPT tracking MP technique  
ACO MPPT track MP based on search space and maintain the MP by the workability of the swarm process  
based on ant movement and pheromone concentration. The convergence of the ACO MPPT sometimes  
becomes premature due to search space unable to identify the nearest local peak power or due to the swarm  
process which converge wrongly at the wrong point [5]. The duty cycles of the search spaces are generated  
randomly at the initial process of the ACO MPPT, this causes the highest power to be generated at this stage is  
declare as local peak power [15]. The local peak power at this stage will determine the convergence of the  
global peak power whether converge at peak power or fall to local minimum.  
Implementing Perturb and Observe MPPT method for PV System  
PNO is the MPPT method that finding MP by observing the PV curve characteristic comparing the old PV  
parameter such as PV voltage, PV current or PV power with the updated PV curve parameters characteristic.  
This research using PV voltage and PV power as perturb technique to identify MP. Fig. 3 shows the PNO  
MPPT technique to track MP.  
Fig. 3 PNO MPPT tracking MP technique.  
PNO MPPT track MP by moving by step to reach MP and maintaining the MP by observing the current MP  
point [3], [4]. The tracking technique of PNO is able to track MP. However, the time taken to track the MP s  
longer compare to the ACO MPPT which takes less time comparing to the PNO MPPT [8]. This is due to by  
step movement taken by PNO MPPT to track MP.  
Combination of ACO and PNO for PV System MPPT  
The reason of ACO and PNO MPPT combination are implemented is to limit the time taken to track the MP as  
well as to allow the MP converge at the right point [9]. The Search space process of the ACO work normally to  
locate the local peak power and the ACO MPPT formula will work with PNO method to locate the MP [10].  
Fig. 4 shows the MPPT from ACO and PNO combination. The idea of combining both MPPT method is to  
allow a more efficient MP tracking process. By joining both MPPT method together, the MP tracking process  
Page 8818  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
takes shorter time to reach MP and able to maintain MP efficiently. Based on Fig. 4, the MPPT process consist  
of search space which enable the MPPT to identify the closes point to the MP. After the Search Space process,  
PNO MPPT take place to identify the highest MP of the PV curves. The PNO MPPT able to maintain the MP  
based on PV power and voltage analysis.  
Fig. 4: The combination of ACO and PNO MPPT.  
The MPPT combination between ACO and PNO MPPT start with initializing  
which is change in  
pheromone concentration. The  
to the MPPT.  
(
, 2); represent random number of pheromone concentration set  
= 1/  
(
, 2);  
and  
is the subtraction of previous and current PV voltage and current. This will ensure that the previous  
PV power and voltage being recorded for MPPT tracking purposes.  
=
=
(
(
, 7) −  
, 3) −  
;
;
ACO MPPT formula take place to measure the amount of pheromone concentration needed to identify the  
amount of pheromone needed for the duty cycle.  
(
)
(
)
(
)
, 1 = 1 − 0.0007 ∗  
− 1,1 +  
;
Pheromone concentration represented by (1-0.007) which is vital for the whole MPPT process. Wrong tuning  
causes the MPPT unable to converge at the MP if implemented ACO MPPT alone without PNO MPPT.  
(
)
Revised concentration of the pheromone represented by  
, 1 , previous revised concentration of  
(
)
pheromone represented by  
− 1,1 and finally  
. After ACO MPPT used to identify  
the amount of pheromone needed, PNO MPPT take place to do perturbation and observation on whether to  
subtract or add the amount of duty cycle.  
< 0  
< 0  
<
(
, 1)  
(
, 1) =  
(
, 1) −  
;
Page 8819  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
>
(
, 1)  
(
, 1) =  
(
, 1) +  
;
< 0  
>
(
, 1)  
(
, 1) =  
(
(
, 1) +  
;
;
<
(
, 1)  
(
, 1) =  
, 1) −  
After PNO MPPT perturbation and observation, the value of the current PV power and voltage is considered as  
old PV Power and Voltage. This process allowed future MPPT PV Power and Voltage comparison between  
new and previous value. The MPPT combination of ACO and PNO need to consider the initialization of  
(
)
, 1 .  
,
,
and  
.
and  
is the randomness value of  
and  
(
)
is randomness for  
, 2 . Fig. 5 below shows the Pseudo Code of the whole coding  
structure.  
// System Control Loop: Continuous MPPT  
IF enable IS TRUE THEN  
// 1. Initialization (Run once)  
INITIALIZE Swarm_Data_Array, Pheromone_Matrix (Tij), iteration = 0.  
SET swarm_size (e.g., number of "ants").  
LOOP forever DO  
IF iteration < swarm_size THEN  
// PHASE 1: Data Collection (Ant Movement)  
1. COLLECT Vpv, Ipv, and Duty_Cycle.  
2. STORE Power_P in the Swarm_Data_Array  
3. INCREMENT iteration.  
ELSE // iteration >= swarm_size  
// PHASE 2: Optimization and Reset (Pheromone Update)  
// A. Evaluation & Reset  
1. FIND and STORE Local_Best_Power (P_max) and its Duty_Cycle from the array.  
2. SET iteration = 1.  
3. SET initialize = 1. // Trigger the optimization block  
IF initialize == 1 THEN  
// B. Core Optimization Block (Swarm Step)  
// 1. Pheromone Update (ACO Global Guidance)  
// Use the best power found to reinforce the optimal path.  
Tij(new) = (rho * Tij(old)) + Delta_Tij // ACO Formula  
// 2. Local Refinement (PNO)  
// Use PNO to precisely tune the new Duty_Cycle suggested by ACO.  
PERFORM PNO_MPPT_CYCLE.  
// 3. Update Records  
UPDATE Local_Best_Power and Duty_Cycle with the new PNO result.  
INCREMENT iteration. // Moves to the next particle's step  
END IF  
// C. Global Best Identification (After one full cycle)  
Page 8820  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
Fig. 5 Shows the Pseudo Code of the hybrid ACO and PNO MPPT method.  
RESULT AND DISCUSSION  
PV Power collected based on constant irradiance value.  
Based on the study and observation on the propose MPPT method. Author conducted an experiment to verify  
the workability of the PV power of the propose MPPT method during dynamic irradiance changes and normal  
circumstances irradiance changes. The graph collected is of type PV power against time changes.  
2
Fig. 6 Comparison of PV Power collected between proposed, ACO and PNO MPPT at 1000  
taken between (0-0.2) and (0.198-0.2) second.  
/
at time  
The workability of the propose MPPT method be able to converge better compare to the normal ACO MPPT  
method. The irradiance value tested is from 100-1000  
2. Based on Table II, the calculated PV power is  
/
use to compare with the actual PV power collected from the analysis.  
2
Fig. 7 Comparison of PV Power collected between ACO & PNO, ACO and PNO MPPT at 1000  
time taken between ((0.199-0.2) second.  
/
at  
Based on Fig. 6 and Fig. 7 it is clear that the proposed MPPT method converge better compare to the ACO and  
PNO MPPT. Fig. 7 shows that the ACO MPPT converge near the MP while PNO and proposed MPPT method  
able to converge at MP. Proposed MPPT method able to maintain at the MP better compare to the PNO MPPT.  
Page 8821  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
2
Fig. 8 Comparison of PV Power collected between proposed, ACO and PNO MPPT at 500  
taken between (0-0.2) and (0.198-0.2) second.  
/
at time  
PNO MPPT require more time to be able to reach the MP and the convergence is sometimes slip due to perturb  
and observe character of the PNO MPPT. Fig. 8 and Fig. 9 show the characteristic of the PV power track at the  
2
irradiance of 500  
/
.
2
Fig. 9 Comparison of PV Power collected between proposed, ACO and PNO MPPT at 500  
taken between (0.199-0.2) second.  
/
at time  
2. The  
Fig. 8 and Fig. 9 show the characteristic of the PV power collected at the irradiance of 500  
/
convergence of the 3 methods able to converge at the MP. However, the proposed method able to converge  
better compare to the ACO and PNO MPPT method. Table IV provide the detail of the average PV power  
based on varying irradiance value.  
TABLE IV The Average Pv Power Data Based on Varying Irradiance.  
Irradiance,  
η
η
η
W/m 2  
1000  
900  
−
−
173.5 173.22  
156.1 155.09  
138.4 137.17  
120.6 119.97  
102.5 101.75  
0.9984  
0.9935  
0.9911  
0.9948  
0.9927  
0.995  
171.02 0.9857 166.98 0.9624  
154.9 0.9923 152.63 0.9778  
135.6 0.9798 132.88 0.9599  
119.16 0.9881 118.55 0.983  
101.94 0.9945 99.27 0.9685  
83.22 0.9872 82.21 0.9752  
800  
700  
600  
500  
84.3  
83.88  
400  
65.95 65.41  
0.9918  
64.97 0.9851 63.6  
0.9644  
Page 8822  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
300  
47.47 47.15  
28.64 28.47  
0.9932  
0.9941  
0.9953  
0.993  
47.08 0.9918 46.48 0.9791  
28.53 0.9962 27.55 0.962  
200  
100  
9.77  
9.72  
9.61  
0.9834 9.59  
0.9884  
0.9815  
0.9714  
Average  
The experimental results indicate that the hybrid ACO-PNO MPPT (Maximum Power Point Tracking) method  
consistently achieves power outputs that are close to the calculated maximum power across various irradiance  
levels. Specifically, it produces power outputs ranging from 9.77 W at 100 W/m² to 173.5 W at 1000 W/m².  
This hybrid approach reaches peak efficiencies of 99.53% at the lowest irradiance level and maintains  
efficiency above 99% even at higher irradiance levels [9]. This demonstrates its superior tracking capability  
compared to the individual ACO and PNO methods [1]. Both standalone algorithms show slightly lower power  
extraction, which suggests that the hybrid algorithm successfully combines the strengths of global search  
(ACO) and local refinement (PNO) to optimize power harvesting. The high efficiency stability across varying  
environmental conditions indicates that this hybrid MPPT strategy is robust and well-suited for practical  
photovoltaic (PV) applications, providing reliable performance even in low-light situations [10]. Overall, the  
data highlights the potential advantages of hybrid methodologies in maximizing PV system efficiency by  
closely approximating the ideal maximum power output.  
The core finding is the hybrid method's exceptional efficiency and accuracy across the full operational range.  
As shown in Table IV and figures like Fig. 6, the hybrid ACO-PNO consistently maintains efficiencies  
exceeding 99.00%, reaching a peak of 99.53% at 100 W/m². This high level of sustained performance across  
varying irradiance, from low-level 100 W/m² to high-level 1000 W/m² scenarios, suggests a robust tracking  
mechanism. This robustness confirms the potential advantages of hybrid methodologies in maximizing PV  
system efficiency, as previously suggested by studies on hybrid intelligent control [16] and metaheuristic  
optimization [6]. The ability to precisely approximate the calculated maximum power is essential for real-  
world applications, supporting its suitability for systems requiring reliable power extraction, such as off-grid  
smart infrastructure [17] or photovoltaic water pumping systems [18]. Hybrid ACO-PNO excels at 99.30%  
efficiency, capturing maximum power, while pure ACO performs strongly at 98.84% with global search,  
occasionally missing the peak; pure PNO trails at 97.14% due to oscillations reducing power extraction.  
PV Power collected during Dynamic irradiance changes  
Dynamic irradiance is use to test the convergence capability of the three method MPPT during instantaneous  
irradiance changes. For the 1 test the irradiance use is 600, 700, 900 and 700  
/
2. The instant irradiance  
changes taking 0.1 second. The test is used to identify the working MPPT at dynamic situation. In the real  
situation, the irradiance changes do not change at instant.  
TABLE V The Dynamic Irradiance Pattern 1.  
Varying Irradiance  
Pattern 1  
Performance  
Parameters  
MPPT  
Average  
Power  
101.754 119.973 155.086 119.973  
Irradiance  
600  
700  
900  
700  
PSO-  
INC  
0.1-0.2 0.2-0.3  
s
0-0.1s  
0.3-0.4 s  
Convergence  
Time  
s
0.02217 0.12526 0.26977 0.324252  
Page 8823  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
This method is used to show that the proposed MPPT able to converge efficiently under dynamic irradiance  
changes. Fig. 10 to Fig. 14 show the analysis of PV Power collected for the 1 dynamic irradiance test. The  
performance analysis of the Ant Colony Optimization and Perturb and Observe (ACO-PNO) MPPT method  
under Pattern 1 shows promising potential for efficient energy harvesting in photovoltaic (PV) systems across  
varying irradiance levels.  
Fig. 10 The analysis on PV power collected for 1 dynamic irradiance changes at time taken between (0-0.4)  
second.  
The average power output increases significantly with irradiation intensity, rising from approximately 101.75  
W at 600 W/m² to about 155.09 W at 900 W/m². This demonstrates that the hybrid MPPT adapts effectively to  
different sunlight conditions and optimizes power extraction.  
Fig. 11 The analysis on PV power collected for 1 dynamic irradiance changes at time taken between (0.08-  
0.1) second.  
Notably, the convergence time increases as irradiance levels rise, ranging from around 0.022 seconds at the  
lowest irradiance to 0.324 seconds at the highest.  
Fig. 12 The analysis on PV power collected for 1 dynamic irradiance changes at time taken between (0.18-  
0.2) second.  
Page 8824  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
This reflects the greater computational effort needed to accurately track the maximum power point (MPP)  
when the irradiance is higher. The initial rapid convergence observed at 600 W/m² within just 0.022 seconds  
indicates that the hybrid approach effectively identifies the MPP under lower irradiance conditions.  
Fig. 13 The analysis on PV power collected for 1 dynamic irradiance changes at time taken between (0.28-  
0.3) second.  
In contrast, the longer convergence times at higher irradiance highlight a trade-off: achieving optimal accuracy  
requires more processing time. Overall, the ACO-PNO pattern demonstrates strong adaptability, effectively  
balancing power maximization with convergence speed, which is crucial for dynamic PV environments.  
Fig. 14 The analysis on PV power collected for 1 dynamic irradiance changes at time taken between (0.38-  
0.4) second.  
Fig. 10 shows the overall convergence of the PV power. During the time between 0.2-0.3 second, ACO and  
PNO MPPT having a trouble to converge at the MP. PNO MPPT need sometimes to use perturb and observe  
method to reach the MP and ACO MPPT need to swarm to reach MP.  
TABLE VI The Dynamic Irradiance Pattern 2.  
Varying Irradiance  
MPPT Performance Parameters  
Pattern 2  
Average Power  
Irradiance  
65.945  
400  
101.754  
600  
93.445  
550  
84.936  
500  
PSO-  
INC  
0-0.1s  
0.01623  
0.1-0.2s  
0.17432  
0.2-0.3s 0.3-0.4s  
0.20712 0.3112  
Convergence Time  
The proposed method able to reach peak power at shorter time due to the capability to swarm and do perturb  
Page 8825  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
and observe at the given time. Fig. 11 to Fig. 14 shows convergence of every MPPT method at closer time and  
PV power range.  
Fig. 15 The analysis on PV power collected for 2 dynamic irradiance changes at time taken between (0-0.4)  
second.  
The convergence ability of proposed MPPT method converge better. For the 2 test the irradiance use is 400,  
600, 550 and 500  
/
2. Fig. 15 to Fig. 19 show the analysis of PV Power collected for the 2 dynamic  
irradiance test. The performance of the ACO-PNO MPPT under Pattern 2 demonstrates a responsive  
adaptation to varying irradiance levels.  
Fig. 16 The analysis on PV power collected for 2 dynamic irradiance changes at time taken between (0.08-  
0.1) second.  
The average power outputs recorded are approximately 65.95 W at 400 W/m², 101.75 W at 600 W/m², 93.45  
W at 550 W/m², and 84.94 W at 500 W/m². This data shows that the hybrid MPPT effectively captures and  
utilizes different sunlight intensities, with power outputs closely aligning with the corresponding irradiance  
levels.  
Fig. 17 The analysis on PV power collected for 2 dynamic irradiance changes at time taken between (0.18-  
0.2) second.  
Page 8826  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
Notably, the convergence times range from very quick (around 0.016 seconds at 400 W/m²) to slower values  
(up to 0.311 seconds at 500 W/m²). This reflects the adaptive complexity of the algorithm in response to  
different environmental conditions. The rapid initial convergence at lower irradiance (0.016 seconds) suggests  
efficient tracking under less intense sunlight, while the longer convergence times at medium irradiance levels  
indicate that more iterations are needed to fine-tune the maximum power point under moderate conditions.  
Fig. 18 The analysis on PV power collected for 2 dynamic irradiance changes at time taken between (0.28-  
0.3) second.  
Overall, Pattern 2 shows that the hybrid ACO-PNO MPPT can reliably maximize power extraction with an  
acceptable convergence speed. This makes it suitable for diverse and fluctuating irradiance scenarios in  
photovoltaic (PV) systems. The results emphasize a good balance between power maximization and  
computational effort, which is crucial for real-time applications.  
Fig. 19 The analysis on PV power collected for 2 dynamic irradiance changes at time taken between (0.38-  
0.4) second.  
Fig. 15 shows the overall convergence of the PV power. During the time between 0.1-0.2 and 0.2-0.3 second,  
the ACO MPPT convergence becomes premature due to instantaneous dynamic irradiance changes and PNO  
MPPT is continue to do perturb and observe method and able to climb near MP at the time spawn between  
0.287-0.3 second. Fig. 16 to Fig. 19 shows the convergence of every MPPT method at close time and PV  
power range. The convergence of the proposed MPPT method able to converge better compare to the ACO and  
PNO MPPT method. The superior dynamic performance of the hybrid ACO-PNO is the most significant result  
of this study. The experiments involving dynamic irradiance changes (Patterns 1 and 2, shown in Tables V and  
VI, and figures like Fig. 10 and Fig. 15) reveal that the combination effectively mitigates the known  
weaknesses of the individual algorithms. Table VII shows the detail information of settling time, oscillation  
and tracking error under rapid changing conditions.  
TABLE VII Settling Time, Oscillation, And Tracking Error Under Rapidly Changing Conditions  
Metric  
P&O (Standard)  
ACO (Standard)  
ACO−PNO (Hybrid)  
Slow-Medium (Depends Medium  
(Fast Fastest (Rapid ACO search + quick  
Settling Time  
Page 8827  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
(to reach 98% heavily on perturbation initial convergence, PNO lock)  
MPP)  
step size)  
but  
slow  
final  
refinement)  
Low-Medium  
(Mandatory (Oscillates around  
oscillation around the peak) discrete pheromone  
path nodes)  
Very Low (close to 0) (PNO is  
used only for refinement, then stops  
when the Delta P condition is met)  
Steady-State  
Oscillation  
Amplitude  
High  
Average  
Tracking  
Error ($\Delta  
P$)  
1% - 3% (Error  
from discretization  
of search space)  
1%  
-
4% (Due to  
0.2% - 1% (Best-in-class, due to  
constant refinement)  
continuous oscillation)  
Handling  
Partial  
Shading  
Poor (Stuck on local Excellent (Designed Excellent (Retains ACO's global  
maxima) for global search) search capability)  
The ACO-PNO hybrid MPPT method excels in dynamic performance by leveraging the strengths of both  
algorithms: it achieves the fastest settling time because the global search of ACO quickly jumps near the new  
Maximum Power Point (MPP), allowing the local search of PNO to take over for rapid final convergence. This  
combination drastically reduces the Average Tracking Error (0.2% - 1%) and Steady-State Oscillation  
Amplitude (close to 0) because, unlike standard PNO which must always oscillate, the refined PNO stage can  
be programmed to stop perturbing and lock the duty cycle once the change in power (Delta P) is below a  
minimum threshold, ensuring superior power extraction stability and efficiency. The standalone P&O method,  
while simple, suffers from a slow, step-by-step movement, leading to longer convergence times and power  
loss, particularly when instantaneous changes occur. Conversely, the standalone ACO MPPT, a bio-inspired  
optimization technique similar to Kinetic Gas Molecular Optimization [19] or Beluga Whale Optimization  
[20], is designed for global searching but demonstrates a tendency for premature convergence during the  
transient periods (as seen in Fig. 15 between 0.1s and 0.3s. This premature convergence means the swarm  
process fails to correctly identify the new global peak, a common challenge in nature-inspired methods. Other  
metaheuristic methods, such as the Horse Herd Optimization Algorithm [21] or Sooty Tern Optimization [22],  
similarly wrestle with the trade-off between speed and local optima trapping. The proposed hybrid ACO-PNO  
acts as a highly effective two-stage MPPT technique [23]. The initial swarming (ACO) quickly guides the  
system toward the general vicinity of the maximum power point (MPP), significantly reducing the required  
tracking time. The subsequent P&O refinement then ensures precise convergence directly onto the MPP,  
correcting the ACO's tendency to settle at a slightly lower, non-optimal point. The resulting  
CONCLUSION  
This analysis demonstrates that the hybrid ACO-PNO MPPT (Maximum Power Point Tracking) method offers  
a promising solution for optimizing the performance of photovoltaic (PV) systems. It achieves high power  
extraction efficiency and closely approximates the ideal power output. Its consistent performance across  
various irradiance levels highlights its suitability for real-world PV applications, where environmental  
conditions can vary significantly. Further research could investigate dynamic parameter tuning to enhance  
performance in low-light conditions and explore the integration of this method into larger PV arrays for  
scalability validation. In conclusion, it is possible to track the global maximum power at each irradiance level  
by adjusting the Particle Swarm Optimization (PSO) output signal distribution characteristics. The speed and  
accuracy of the power tracking process can be modified by adjusting the inertia weight, swarm size, and  
individual learning rate in the PSO algorithm. Rapid tracking of maximum power is crucial for identifying  
power changes within the PV system. Through simulation software, power variations can be observed from the  
PSO algorithm after modifications, ensuring optimal maximum power point tracking. The characteristics of PV  
voltage and current are influenced by the behavior of the PWM (Pulse Width Modulation) signal.  
Page 8828  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
ACKNOWLEDGMENT  
For the opportunity to participate in this research, the authors would like to thank Center for Advanced  
Computing Technology (C-ACT), Fakulti Kecerdasan Buatan dan Keselamatan Siber (FAIX), Fakulti  
Teknologi Maklumat dan Komunikasi (FTMK) and Centre for Research and Innovation Management (CRIM),  
Universiti Teknikal Malaysia Melaka (UTeM) for providing the facilities and support for this research work.  
We'd also like to express our gratitude to the UTeM's Financial Support for funding the project.  
REFERENCES  
1. M. H. Ali, M. Zakaria, and S. El-Tawab, “A comprehensive study of recent maximum power  
point tracking techniques for photovoltaic systems,” Sci Rep, vol. 15, no. 1, Dec. 2025, doi:  
10.1038/s41598-025-96247-5.  
2. A. Elsafi, A. A. Almohammedi, M. Balfaqih, Z. Balfagih, and S. Sabri, “Comparative analysis  
of maximum power point tracking methods for power optimization in grid tied photovoltaic  
solar systems,” Discover Applied Sciences, vol. 7, no. 9, Sep. 2025, doi: 10.1007/s42452-025-  
07606-w.  
3. A. B. Djilali et al., “Enhanced variable step sizes perturb and observe MPPT control to reduce  
energy loss in photovoltaic systems,” Sci Rep, vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-  
025-95309-y.  
4. D. Sibtain, M. M. Gulzar, K. Shahid, I. Javed, S. Murawwat, and M. M. Hussain, “Stability  
Analysis and Design of Variable Step-Size P&O Algorithm Based on Fuzzy Robust Tracking  
of MPPT for Standalone/Grid Connected Power System,” Sustainability (Switzerland), vol. 14,  
no. 15, Aug. 2022, doi: 10.3390/su14158986.  
5. B. Babes, A. Boutaghane, and N. Hamouda, “A novel nature-inspired maximum power point  
tracking (MPPT) controller based on ACO-ANN algorithm for photovoltaic (PV) system fed  
arc welding machines,” Neural Comput Appl, vol. 34, no. 1, pp. 299–317, Jan. 2022, doi:  
10.1007/s00521-021-06393-w.  
6. M. Yaich, Y. Dhieb, M. Bouzguenda, and M. Ghariani, “Metaheuristic Optimization Algorithm  
of MPPT Controller for PV system application,” in E3S Web of Conferences, EDP Sciences,  
Jan. 2022. doi: 10.1051/e3sconf/202233600036.  
7. N. Priyadarshi, V. K. Ramachandaramurthy, S. Padmanaban, and F. Azam, “An ant colony  
optimized mppt for standalone hybrid pv-wind power system with single cuk converter,”  
Energies (Basel), vol. 12, no. 1, Jan. 2019, doi: 10.3390/en12010167.  
8. M. Sedraoui et al., “Development of a fixed-order controller for a robust P&O-MPPT strategy  
to control poly-crystalline solar PV energy systems,” Sci Rep, vol. 15, no. 1, Dec. 2025, doi:  
10.1038/s41598-025-86477-y.  
9. B. Naima et al., “Enhancing MPPT optimization with hybrid predictive control and adaptive  
P&O for better efficiency and power quality in PV systems,” Sci Rep, vol. 15, no. 1, Dec.  
2025, doi: 10.1038/s41598-025-10335-0.  
10. [K. H. Chao and M. N. Rizal, “A hybrid mppt controller based on the genetic algorithm and ant  
colony optimization for photovoltaic systems under partially shaded conditions,” Energies  
(Basel), vol. 14, no. 10, May 2021, doi: 10.3390/en14102902.  
11. K.Burhanudin, N.A.Kamarzaman, A.A.A.Samat, A.I.Tajudin, S.S.Ramli, and N.Hidayat,  
Implementing Boost Converter Algorithm with PSO for Photovoltaic System During Partial  
Shading Condition. Johor Bharu: IEEE, 2015. doi: 10.1109/CENCON.2015.7409576.  
12. A. Nasir, I. Rasool, D. Sibtain, and R. Kamran, “Adaptive Fractional Order PID Controller  
Based MPPT for PV Connected Grid System Under Changing Weather Conditions,” Journal of  
Electrical Engineering and Technology, vol. 16, no. 5, pp. 2599–2610, Sep. 2021, doi:  
10.1007/s42835-021-00782-w.  
13. X. Liu et al., “Simulation-Based Evaluation of Power Efficiency and Output Capacitance in  
Standalone PV MPPT Buck Converters Using 200 V p-GaN HEMTs,” Journal of Electrical  
Engineering and Technology, vol. 20, no. 6, pp. 3875–3887, Sep. 2025, doi: 10.1007/s42835-  
025-02334-y.  
Page 8829  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025  
14. V. C. Kotak, P. Tyagi, and A. Professor, “DC To DC Converter in Maximum Power Point  
Tracker,” International Journal of Advanced Research in Electrical, Electronics and  
Instrumentation Engineering (An ISO Certified Organization), vol. 3297, no. 12, pp. 6115–  
6125, 2007.  
15. M. Q. Taha, M. K. Mohammed, and B. El Heiba, “Metaheuristic Optimization of Maximum  
Power Point Tracking in PV Array under Partial Shading,” Engineering, Technology and  
Applied Science Research, vol. 14, no. 3, pp. 14628–14633, Jun. 2024, doi:  
10.48084/etasr.7385.  
16. H. Azizi-Monfared and M. Ghanbarisabagh, “Hybrid Intelligent Control and Maximum Power  
Point Tracking of a Solar Generator under Variable Irradiance and Temperature using a Multi-  
method Approach,” International Journal of Engineering, Transactions B: Applications, vol. 39,  
no. 6, pp. 1462–1481, Jun. 2026, doi: 10.5829/ije.2026.39.06c.14.  
17. S. Vadi, “Design and Implementation of an Off-Grid Smart Street Lighting System Using  
LoRaWAN and Hybrid Renewable Energy for Energy-Efficient Urban Infrastructure,” Sensors,  
vol. 25, no. 17, Sep. 2025, doi: 10.3390/s25175579.  
18. F. Id Ouissaaden, H. Kamel, and S. Dlimi, “Simulation and Performance Evaluation of a  
Photovoltaic Water Pumping System with Hybrid Maximum Power Point Technique (MPPT)  
for Remote Rural Areas,” Processes, vol. 13, no. 9, Sep. 2025, doi: 10.3390/pr13092867.  
19. R. S. Prasad and R. Thiyagarajan, “Kinetic gas molecular optimization method for PV system  
under partial shaded condition,” International Journal of Intelligent Engineering and Systems,  
vol. 13, no. 4, pp. 33–43, 2020, doi: 10.22266/IJIES2020.0831.04.  
20. B. Kothapalli and G. T. Sundar Rajan, “Hybrid beluga whale optimization based MPPT for  
photovoltaic powered open end winding induction motor drives,” Sci Rep, vol. 15, no. 1, Oct.  
2025, doi: 10.1038/s41598-025-15680-8.  
21. A. W. Ibrahim et al., “A high-speed MPPT based horse herd optimization algorithm with  
dynamic linear active disturbance rejection control for PV battery charging system,” Sci Rep,  
vol. 15, no. 1, p. 3229, Dec. 2025, doi: 10.1038/s41598-025-85481-6.  
22. M. T. Kaaitan, R. A. Fayadh, Z. S. AL-sagar, S. J. Yaqoob, M. Bajaj, and M. S. Geremew, “A  
novel global MPPT method based on sooty tern optimization for photovoltaic systems under  
complex partial shading,” Sci Rep, vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-025-13007-  
1.  
23. Q. Pang, F. Zhang, S. Han, T. Zhou, and Y. Wang, “A new two-stage MPPT technique for  
enhancing the performance of PV system,” Electrical Engineering, vol. 107, no. 8, pp. 10899–  
10909, Aug. 2025, doi: 10.1007/s00202-025-03067-x.  
Page 8830  
www.rsisinternational.org