Convolutional Neural Network Based Metal Object Detection System for Wireless EV Charging

Chengyin Liu,Hao Chen,Zeqian Cheng,Yizhen Lin,Jiande Wu,Xiangning He
DOI: https://doi.org/10.1109/ecce50734.2022.9947416
2022-01-01
Abstract:In wireless electric vehicle charging, metal object detection (MOD) is a vital guarantee for the reliable operation of the system. In this paper, a magnetic field distribution feature learning based metal object detection method is proposed. A reusable hardware detection system using tunneling magnetoresistive (TMR) sensor matrix is built to measure the magnetic field distribution. Model learning is used to extract features of magnetic field distribution after constructing a training set based on collection data. convolutional neural network (CNN) based metal object detection model is also derived. The experimental results show that the proposed system detects metal objects with high accuracy, demonstrating the efficacy of the proposed method.
What problem does this paper attempt to address?