Multiobjective Parameter Optimization of a Four-Plate Capacitive Power Transfer System

Abstract

For a four-plate capacitive power transfer (CPT) system, the existing coupling and compensation design methods can only provide an approximate local parameter optimization solution, instead of a globally optimized solution. This article proposes an accurate calculation method for cross-coupling capacitance and a multiobjective optimization algorithm combining nondominated sorting and Kent map for an SS-type CPT system. Mathematical models of the cross-coupling capacitances and port capacitances are established based on a two-port capacitive coupling model of the four plates. Besides, the constraints considering the reasonable region of the parameter values are given. Taking the output power, efficiency, and total harmonic distortion (THD) as objective functions, with the power transfer distance, operating frequency, load resistance, and input voltage as decision variables, the Pareto-optimal fronts are obtained by the proposed algorithm. A group of global optimal parameters is chosen from the solution sets, and a simulation model and a prototype of the CPT system are constructed. The efficiency of 87.1% at 252.3-W output power has been achieved. The simulation and experimental results verified the feasibility and effectiveness of the proposed calculating method for the cross-coupling capacitance and optimization method.