Dynamic reconfiguration method and performance optimization of 3D curved photovoltaic modules under non-uniform illumination conditions based on improved simulated annealing algorithm (ISAA)

Abstract

Three-dimensional curved photovoltaic (PV) modules offer flexible integration on irregular surfaces such as building facades, vehicles, and wearable devices, but it faces severe power losses from non-uniform irradiance and current mismatch. This study develops an Improved Simulated Annealing Algorithm (ISAA) for real-time electrical topology reconfiguration of a 6 × 6 curved PV module. The ISAA reconfiguration method embeds facet-level irradiance feedback into the annealing search, integrates a global-best current-mismatch index into its acceptance criterion to overcome local optima, and executes parallel, parameter-free cell-swap trials for rapid convergence. This paper compares the performance of ISAA, TCT, and SDS reconfiguration methods under various illumination distribution conditions and evaluates the adaptability under dynamic irradiance environments. Results show ISAA maintains a current mismatch index below one in all cases, smooths I–V/P–V characteristics. In addition, ISAA effectively reduces the cumulative effect of current mismatch and maintains relatively stable performance improvement under both sunny and cloudy weather conditions. ISAA reduces daily mismatch losses by up to 52 %, boosts daily energy yield by 42 % ∼ 44 % compared to TCT and SDS, and enhances fill factor and operational stability under rapidly changing irradiance. Overall, ISAA handles complex irradiance, dynamic shading, and weather fluctuations, enabling optimized 3D PV performance.