Modeless Prediction of Variable Coupling Effect for Multiple-Load Wireless Power Transfer

This article proposes a modeless prediction scheme to estimate the coupling effect for wireless power transfer (WPT) systems. To address the impact of the variable coupling effect for WPT systems, the acquirement of the coupling effect has been gradually one of the most important research topics in recent years. However, the mutual inductance cannot be directly measured, because it is determined not only by the distance but also by various external factors. Although there are several efforts for the prediction of the mutual inductance that have been made in previous studies, the practicability of existed schemes mostly suffers from limitations caused by the communication link, the complex calculation, and the measurement accuracy. Accordingly, this article proposes a modeless prediction scheme using neural network (NN) to dynamically estimate the varying coupling effect caused by practical working conditions like the variation of relative position between the transmitting and the pickup coils. In this article, the simulated and experimental results are both given to verify the feasibility of the proposed modeless prediction scheme as well as its salient advantages of being communication free, calculation simple, and measurement convenient.