Optimal C-type filter design for wireless power transfer system by using support vector machines

Abstract

The rapid increase in the number of Electrical Vehicles (EVs) will bring difficulties in the management of charging process and pose serious grid problems at low voltage levels. Particularly, with employment of wireless power transfer (WPT) system in a charging station, harmonic interference will increase. The main reason of that poor power quality lies on high frequency square wave output of transmitter side of WPT. In this study, a support vector machine (SVM) is proposed to design an optimal C-type passive filter in order to mitigate voltage and current total harmonic distortions (THD) of WPT system. Hereby, SVM-based model is constructed which consists of THD indices and power factor (PF) as outputs whereas filter parameters are inputs. The main aim of optimization process is minimization of distortions and correction of PF while searching the filter parameters. Particle swarm optimization (PSO) algorithm is employed to find the optimal filter parameters. To show the efficiency of proposed method, simulation studies are carried out on Matlab®/SimulinkTM environment. It is observed that voltage total harmonic distortion (THDv) and current total harmonic distortion (THDi) are calculated as 1.03%, 2.23%, respectively, and the power factor is improved to 0.995% when the designed C-type filter is utilized.