Machine Learning Approach of Optimal Frequency Tuning for Capacitive Wireless Power Transfer System

Abstract

Wireless power transmission has become a remarkable research topic due to its enormous application potential. Recent advances in machine learning have been shown to be the most promising approach that offers significant capabilities in the wireless power transfer system (WPTS) for selecting the optimal frequency tuning to achieve high efficiency performance. However, developing an automated frequency-tuned system remains a challenge. In this study, a novel frequency-tuned method is presented that utilises machine learning-based models such as neural networks (NN), support vector regression (SVR), and linear regression (LR) to estimate the best efficiency provided by the frequency level at the most optimal frequency tuning level from the experimental dataset, which is capable of aiding in the selection of the most efficient WPTS design. The results show that the SVR has the highest degree of accuracy, making it a promising option for optimising the tuning of power transfer systems while enhancing their performance efficiency. Keywords—Efficiency, frequency tuning, machine learning, neural network, power transfer