Yohanes Leonaldo Sinaga, Elya Syafa'atun Ni'mah, Muhammad Daffa Pratama, Arvina Rizqi Nurul'aini, Mohammad Mahruf Alam, Rizky Ajie Aprilianto
Global energy demand has intensified with industrialization, electrification of transportation, and digital transformation, positioning solar photovoltaic (PV) technology as the fastest-growing renewable source. Nevertheless, the effective performance of PV systems is constrained by irradiance variability, temperature fluctuations, and partial shading conditions, which disrupt maximum power point (MPP) operation. Conventional MPP tracking methods, while simple, suffer from slow convergence and steady-state oscillations, reducing their effectiveness in large-scale gridconnected applications. To address these challenges, recent studies have explored metaheuristic algorithms. This paper provides a systematic review of metaheuristic-based Maximum Power Point Tracking for PV systems, conducted using the population, intervention, comparison, outcome, and context (PICOC) framework and preferred reporting items for systematic reviews and meta-analyses (PRISMA) methodology. A total of 52 articles published between 2021 and 2025 are analysed, focusing on four thresholds, efficiency, convergence time, oscillation, and computational complexity. The results show trends for Research Question (RQ1-RQ2), Particle Swarm Optimizer appears most frequently, followed by Genetic Algorithm. Simulation studies still dominate over hardware validation. For comparative performance, RQ2 highlighted the top 5 algorithms in the review, including PSObased sensorless backstepping with Immersion and Invariance tuning, Marine Predator and Mayfly Optimization duty-cycle search, Growth Optimization Algorithm with Deep Neural Network and Fractional Order Proportional-IntegralDerivative tuning, Reptile Search Algorithm two-stage duty optimization, and Snake Optimizer Algorithm. These top 5 algorithms surpassed the predefined thresholds across the four evaluation dimensions (efficiency, convergence time, oscillation, and algorithmic complexity) and thus merit priority for follow-up hardware validation and deployment. © 2025 IEEE.
Universitas Negeri Semarang, Department of Electrical Engineering, Semarang, Indonesia
