<![CDATA[The growing demand for reliable and cost-effective electricity has accelerated the adoption of solar photovoltaic (PV) systems; however, inappropriate system sizing can lead to excessive installation costs through oversizing or unreliable power supply through undersizing. This study aims to optimize the sizing of a PV power system within the 10–20 kW capacity range using an artificial intelligence-based approach. A detailed PV system model incorporating half-cut monocrystalline PV modules and lithium iron phosphate (LiFePO₄) battery storage was developed in MATLAB/Simulink. Solar irradiance and ambient temperature were incorporated to represent realistic operating conditions. Particle Swarm Optimization (PSO), implemented in Python, was used to determine the optimal PV array configuration and battery capacity. System performance was evaluated based on energy output, Loss of Power Supply Probability (LPSP), and Net Present Cost (NPC). The simulations showed that PV output increased with solar irradiance, whereas module efficiency declined as temperature increased. The optimization identified a configuration of 32 PV modules, comprising eight modules in series and four parallel strings, combined with a battery capacity of approximately 690 Ah. This configuration produced a system capacity of approximately 16 kW, achieved an LPSP of 0.006, and reduced system cost relative to larger configurations. These findings demonstrate that PSO-based optimization can improve the technical and economic performance of PV system sizing by balancing energy reliability, storage capacity, and system cost. The study provides an integrated optimization framework for designing reliable and cost-efficient PV–battery systems under variable environmental conditions.]]>
Copyrights © 2026
