Research on coil identification algorithm of wireless power transmission system based on deep learning

Abstract

In the practical application of wireless power transfer (WPT) technology, the identification of uncertain secondary side coil is of great significance to the interoperability of the WPT system. Whether the secondary coil can be accurately identified to realize the ideal match of the primary coil and the secondary coil is one of the keys to achieve efficient power transmission. Based on the characteristics of magnetic field image, this paper combines the convolutional neural network (CNN) to realize the recognition of the coil type, and establishes a 4-layer and 6-layer neural network respectively, and the recognition accuracy is as high as 98%. Compared with the existing visual geometry group (VGG) 16 neural network algorithm, the proposed algorithm has lower complexity and is more suitable for the WPT application field.