Coil optimization and power orientation strategy of wireless power transfer system based on array coils

Abstract

Aiming at the problem of power reduction when the transmitter and receiver coils are misaligned in wireless power transfer (WPT) system, a strategy to improve power by controlling the current phase of the array coils is proposed in this paper. Taking the square coil as an example, the expression of the magnetic field strength at a certain point in the array coil is discussed, and the parameters affecting the magnetic field distribution are determined. In order to improve the uniformity of the magnetic field distribution, the non-equidistant coil structure of the array coil is designed. BP neural network is introduced to learn the variation law of the magnetic flux with the current phase of the array coil in the specified area, and the phase difference between the coils when the magnetic flux is maximum is determined. Finally, the above concepts are realized through simulations and experiments. The results show that the proposed non-equidistant coil has more advantages in the uniformity of magnetic field distribution than the equidistant coil. Under the optimal phase configuration, the magnetic flux in the planar area of the array coil can be effectively improved.

Keywords

• Wireless power transfer (WPT)
• Deep learning
• Convolutional neural networks (CNN)
• Identification algorithm