Switched Ring Particle Swarm Optimizer for Trace of Time-Variant Maximum Power Point

This paper presents a switched ring particle swarm optimizer and considers its application to trace of the maximum power point in time-variant photovoltaic systems. In order to avoid trapping into local optima, the optimizer uses the local best of directed ring topology where the direction is switched periodically. In order to follow the time-variant characteristics, the optimizer uses a flexible renewal method of the personal best. The optimizer has only one particle and sampled values of the particle in an interval are used as plural particles. The cost function of the voltage can be measured only at one point at an instant in real environment. Performing basic numerical experiments for a typical example, the algorithm efficiency is confirmed.