Photovoltaic Maximum Power Point Tracking with Improved Differential Evolution

To address the problem of slow tracking speed of existing differential evolutionary algorithm in the case of local shading, a new maximum power point tracking(abbr.MPPT) control strategy for photovoltaic(abbr.PV) power generation system with improved differential evolutionary algorithm was proposed. Firstly, the triangular mutation strategy was introduced and embedded in the crossover formula of the traditional differential evolutionary algorithm to avoid duplicate sampled power due to the same duty cycle of the output before and after the iteration. Secondly, an adaptive scaling factor strategy was introduced to enhance individual convergence. Finally, the global exploitation and local exploration capabilities of the proposed algorithm were balanced by a reward-penalty mechanism. The simulation results show that the improved algorithm has significant advantages in tracking speed and stability compared with the differential evolution algorithm of the comparative literature.