Abstract: Under partial shaded conditions, the P-U characteristic curve of the PV array will present multi-peak phenomenon. In this case, the traditional maximum power point tracking（MPPT）tends to become invalid due to local optimization and fails to find the maximum power point. To solve this problem, an advanced MPPT method based on improved particle swarm optimization（PSO）algorithm is proposed in this paper. By analyzing the output characteristic curve of the photovoltaic array under partial shaded conditions,the peaks of possible voltage is calculated and used as the initial value of the particle position. The inertia weight and learning factors are adjusted through introduced the state factor to avoid particles falling into local optimum so as to reduce the oscillation of the tracking process. The superiority of the improved PSO algorithm is verified by MATLAB simulation results by comparing the traditional PSO algorithm with the improved PSO algorithm.