n this paper, a maximum power point tracking (MPPT) method that incorporated shading and failure conditions in photovoltaic (PV) module arrays is developed. This MPPT method was built using improved particle swarm optimization (PSO). The PSO algorithm enables PV module arrays to perform MPPT for multi-peak power–voltage (P–V) output characteristic curves when shading or failures occur. This facilitates the tracking of actual maximum power points in PV module arrays. The HIP 2717 PV module produced by SANYO Electric Co., Ltd. was used in this study to assemble various array configurations. The characteristic curves of these array configurations when partial module shading or failure occurred were investigated. Numerous working conditions were selected for dual-peak, three-peak, and four-peak characteristics. PIC microcontrollers were then used to apply both the traditional and the proposed PSO algorithms to enable MPPT. A comparison of the measurement results showed that the proposed PSO algorithm exhibited superior tracking speed, response, and accuracy, compared with those of the traditional PSO algorithm.