Metal Foreign Object Detection In Wireless Charging System Based On Machine Vision And Auxiliary Coil

In recent years,major automobile consumer countries have begun to vigorously develop electric vehicles,but defects such as bulky charging heads and easy plugging and pulling have caused a lot of inconvenience to applications.Now more and more researchers and companies are turning their attention to wireless charging,because of its non-contact advantages to solve the problems caused by plugging and unplugging of the charging head;However,on the other hand,due to the primary device and vice of the wireless charging system The presence of gaps between the side devices makes it easier for metal foreign objects to fall into it,the charging efficiency caused by the metal foreign objects decreases,and heat generation easily causes fires and other problems,which greatly hinder the promotion of wireless charging.This paper comprehensively considers the whole process of reverse storage and parking charging and the characteristics of each stage,and proposes a detection scheme based on machine vision and auxiliary coils to ensure the energy transmission efficiency and charging safety of the charging process.Firstly,it summarizes the research status of metal foreign body detection at home and abroad,summarizes the advantages and disadvantages of each detection scheme,and gives a detailed description of the scheme in this article.Based on the SS-type topology,a mathematical model of wireless energy transmission is established,and the metal foreign objects are divided into non-ferromagnetic metal foreign objects and ferromagnetic metal foreign objects,and equivalent models are established respectively.The effects of metal foreign objects on circuit parameters are theoretically analyzed.Then use the visual detection scheme in the reverse storage phase to extract the wireless charging area according to the corresponding feature design algorithm,and perform spatial correction on the extracted wireless charging area,and collect a large number of corrected charging area images as sample data sets The metal foreign objects in are marked as the category to be detected.A deep learning detection network framework is designed according to the characteristics of metal foreign bodies.Then,during parking charging,the auxiliary coil is designed using the principle of electromagnetic induction to detect metal foreign objects.The detection principle of the symmetric auxiliary coil structure is introduced,and the primary side detection coil and the secondary side detection coil are used to jointly locate the position of the metal foreign body.Based on OP37,the amplifying circuit and filter circuit of this paper are designed and verified by simulation.Finally,an experimental platform was built to test the effects of metal foreign object detection based on vision and auxiliary coil.The parameters were debugged and optimized according to the experimental results,which verified the feasibility of the metal foreign object detection scheme in this paper.