Multi-objective particle swarm optimization applied to the design of Wireless Power Transfer systems

Abstract

This paper proposes a stochastic method - particle swarm optimization (PSO) and Pareto front technique, to conduct a multivariable optimization and design an inductively coupled power transfer system. Previously, design methods have been proposed which require designer experience; they are not only computationally challenging, but also frequently result in suboptimum solutions. The algorithm proposed in this paper models all losses and inductance matrix analytically. We study four common compensation structures, and select a series-parallel topology to design a 200 W experimental prototype optimized with respect to transfer efficiency and VA rating. Experiments and numerical simulations are employed to verify the optimization algorithm.