Maximum Power Point Tracking of Photovoltaic Array on a USV: A Fuzzy Neural-Directed Adaptive Particle Swarm Optimization Approach

Photovoltaic (PV) array equipped on an unmanned surface vehicle (USV) suffers from rapid-changing partial-shading conditions since USV maneuvers frequently alter shadows on deck, thereby facing a challenge in time-varying maximum power point tracking (MPPT). In this paper, a fuzzy neural directed adaptive particle optimization (FN-APSO) solution is innovatively provided to dynamically determine the global maximum power point (GMPP) in a fast-accurate manner. To facilitate the accuracy, an adaptive PSO (APSO) algorithm is created by assigning region-wise update laws which sufficiently avoid unnecessary search behaviors and ensure global convergence, simultaneously. To further enhance the rapidity, using history data, a fuzzy neural network is devised to judge the evolution direction of GMPP, and enables the APSO to incrementally execute, thereby establishing the entire FN-APSO scheme. Simulation results clearly show remarkable MPPT performance in terms of both speed and accuracy under rapid-changing partial-shading conditions.