The maximum power point tracking algorithm is one of the most effective ways to maximize the utilization of solar energy. The non-linearity of PV modules and the multi-peak characteristics of PV arrays make it more difficult to track the maximum power points. Traditional algorithms are easy to fall into local extremes. Most intelligent algorithms have problems such as high computational cost, long time and system oscillation, which will lead to the instability of the system. The radial movement optimization algorithm has the advantages of simplicity, fastness and elimination of oscillation. However, in the search process, the randomness of the particles leads to the oscillation of the system. In view of this problem, an improved method is proposed. Finally, the PV model based on the BOOST circuit is established by AD260. we test the PV cells MPPT under each light and MPPT of the PV array under local shadows. The results show that the improved algorithm can be fast Accurately perform maximum power point tracking.