基于径向基神经网络的多负载无线电能传输系统自适应阻抗匹配方法

An Adaptive Impedance Matching Method Based on Radial Basis Function Neural Network in Multi-Load Wireless Power Transfer Systems

随着无线电能传输(WPT)技术的发展,同时给多个负载供电受到越来越多的关注。多负载WPT系统中,接收线圈之间的交叉耦合会使系统失谐,导致谐振频率处传输效率下降。该文首先分析了交叉耦合对传输效率的影响机理;然后提出基于径向基(RBF)神经网络算法的“T”型阻抗匹配网络,可根据不同负载对匹配网络中电容进行实时调整,实现系统与负载的自适应匹配;最后针对该方法搭建了实验平台,实验结果表明在交叉耦合影响最大时,系统传输效率从最低时的34%提升到了78%。

With the development of wireless power transfer(WPT)technology,more and more attention has been paid to supplying electricity to multiple loads simultaneously.In a multi-load WPT system,the cross coupling between the receiving coils will detune the system,resulting in a decrease of transmission efficiency at the resonant frequency.This paper firstly analyzes the influence of cross coupling on transmission efficiency,and proposes a"T"type impedance matching network based on RBF neural network.The proposed method adjusts the capacitance in the matching network in real time according to different loads,realizing the adaptive matching of system and load.Finally,an experimental setup was built for the proposed method,and the results show that when the cross coupling is the greatest,the system transmission efficiency increased from 34%to 78%.