Particle Swarm Optimization with Reducing Boundaries (PSO-RB) for Maximum Power Point Tracking of Partially Shaded PV Arrays

TLDR: In this article, a modified PSO algorithm with dynamically reduced boundary (PSO-RB) was proposed for the search space of the particles, which showed that the global maxima search time can be significantly improved while maintaining the accuracy.

Abstract: Multiple Local Maximums (LM) and one Global Maximum (GM) can be present on the Power-Voltage (P-V) output curve during partial shading conditions of a Photovoltaic system. Particle Swarm Optimization (PSO) can consistently find the GM 100% of the time when the swarm size is large enough but the time to find the GM also increases as the swarm size increases - thus reducing the speed of tracking. In this work, we present the performance results of a modified PSO algorithm with dynamically reduced boundary (PSO-RB) for the search space of the particles. Our results indicate that the global maxima search time can be significantly improved while maintaining the accuracy by implementing the PSO-RB algorithm. A direct comparison between the original PSO algorithm and the new PSO-RB algorithm showed that PSO-RB can find the GM point 54.3% (over 2 times) faster than the conventional PSO.