Under partial shading conditions, the output power of photovoltaic arrays has multiple peaks. The traditional maximum power point tracking (MPPT) technology is easy to fall into local peaks during the optimization process, and it is difficult to quickly and accurately track the maximum power point. To solve this problem, an improved sparrow search algorithm is proposed to apply to photovoltaic MPPT technology. The flight step is introduced based on the crow search algorithm, and a dynamic gradient rule is used to adjust the flight step to enhance the exploration ability of the participants, and solve the problem of low optimization accuracy of the sparrow search algorithm in low dimensions. A selfadaption rule is designed and applied to the update of the finder′s position, and the sparrow′s position update is constrained by the boundary processing strategy, which ensures that the sparrow search algorithm can effectively solve the optimization problem in the case of a small search range. The six benchmark functions are tested and verified in MATLAB, and the improved algorithm is applied to the maximum output power point tracking of photovoltaic arrays under uneven illumination conditions for the simulation. The results show that the proposed improved sparrow search algorithm has fast convergence speed and high tracking accuracy, and can effectively avoid falling into local peaks and track the maximum power point quickly and accurately during the optimization process.