Abstract: Particle swarm optimization (PSO) is a kind of intelligent algorithm and has been used in MPPT for a PV array because it has better global search capability, but the tracking speed of PSO algorithm is slow and unstable convergence. The improved PSO algorithm is proposed in this paper according to PV array output characteristics curve analysis which is under condition of partial shaded. The duty ratio of converter is defined as particle, the particles are uniformly distributed at the possible peaks during initialization.The inertia weight and learning factors are adjusted according to the number of iterations, and the velocity updating formula of traditional PSO algorithm is modified through introduced the inverse tangent function to restrict the speed of each particle separately, so that the maximum power point could be found quickly to improve the convergence speed. Finally the simulation results verify that compared with the conventional PSO algorithm, IPSO algorithm has the advantaged of tracking fast and small fluctuations in the tracking process.