Research on maximum power point tracking of photovoltaic power generation based on improved hybrid optimization algorithm.

Due to environmental factors' influence, the power-voltage (P-V) curve of a photovoltaic array typically presents multiple peaks. The traditional gravitational search algorithm is inclined to fall into local optimal solutions and demonstrates poor performance in maximum power point tracking. Consequently, this paper proposes applying the PSO algorithm for optimizing the GSA parameters. Meanwhile, introducing the gravitational constant into the GSA algorithm accomplishes the dynamic adjustment of the three key parameters in PSO. Furthermore, the Levy flight step is incorporated to enhance the global search capability. The improved algorithm can improve the search speed and accuracy by adding memory and group interaction to the particle update formula to alleviate oscillation. Simulink modeling and simulation analysis reveals that, compared with traditional algorithms, the improved algorithm can identify the maximum power point of the photovoltaic array more rapidly and stably under static and dynamic shading conditions.