Receiver Position Estimation Method for Multitransmitter WPT System Based on Machine Learning

This article proposes a receiver position estimation method suitable for multitransmitter(multi-TX)wireless power transfer systems from a new perspective of machine learning. The novel method uses the voltage and current of the transmitter side, and is realized through self-learning, which can solve the problems of complex electromagnetic environment and slow response in the traditional method. First, the switching criterion of the multi-TX wireless power transfer (WPT) system based on receiver position is established. In addition, the receiver position estimation method is determined by comparing the estimation accuracy of the back propagation (BP) and support vector regression (SVR) algorithms. On this basis, a swithcing control strategy of multi-TX WPT system based on receiver position estimation is further proposed. Finally, the experimental prototype with FPGA controller is designed to verify the correctness of the theoretical analysis. The average relative error of the position estimation method is about 5%, and the system efficiency is stable near 90% with the misalignment conditions.