Under the partial shade condition of photovoltaic (PV) arrays, multiple peaks will appear on the P-U curve, and the traditional particle swarm optimization (PSO) algorithm cannot search for the maximum power point quickly and accurately. In view of this issue, a maximum power point tracking (MPPT) control strategy based on improved PSO algorithm using chaotic mapping and Gaussian perturbation is put forward. First, a chaotic Sine mapping was introduced to construct a nonlinear random incremental inertia weight, and Gaussian perturbation was introduced in the “individual cognition” part of the particle swarm. Meanwhile, a logarithmic function was used to construct a learning factor, thus forming an improved PSO algorithm based on chaotic mapping and Gaussian perturbation. Through the test of six typical unimodal and multimodal functions, it was verified that this algorithm converged faster and was less susceptible to local maxima. Then, the algorithm was applied to MPPT control, and comparative simulations of different algorithms on MPPT control were performed. The comparative simulation results show that in three cases of uniform light intensity, static partial shading and dynamic shading, the MPPT control strategy based on improved PSO algorithm using chaotic mapping and Gaussian perturbation has a faster convergence speed and a smaller search oscillation amplitude, and it can find the maximum power point with a higher searching precision, thereby improving the power generation efficiency of the MPPT system.