Foreign Object Detection Method Of Wireless Charging System Based On Impedance Characteristics And Machine Learning

In the actual use of inductive power transmission technology,there will be a magnetic field coupling area between the primary and secondary coils.Foreign objects can easily enter the surface of the primary coil,seriously affecting the stability of the system,reducing the system efficiency,and even damaging the device,causing a fire.Foreign object detection technology(Foreigner Object Detect,FOD)should therefore be born,and has become an important branch of wireless energy transmission technology.The foreign object detection system needs to quickly detect foreign objects entering the coupled magnetic field to ensure the safety of the charging process and the reliability of the system.The detection principle of the impedance characteristic is based on the change of magnetic field parameters caused by different types of foreign objects entering the magnetic field.The changes of the magnetic field parameters are converted into changes in voltage amplitude and phase difference through the circuit,so the amplitude caused by the foreign objects can be extracted Phase parameter changes are detected.This paper focuses on the key technical issues of foreign object detection systems based on impedance characteristics,such as detection coil design,detection circuit design,data feature selection and extraction,and machine learning model construction.Finally,through experiments,it was verified that this method can quickly detect and classify foreign objects in practical applications.The detection principle of impedance characteristics is based on the change of magnetic field parameters caused by different types of foreign objects entering the magnetic field.The changes of the magnetic field parameters are converted into changes in voltage amplitude and phase difference through the circuit,so the amplitude caused by the foreign objects can be extracted Phase parameter changes are detected.The main work and research content of the paper are:(1)Principle analysis of foreign object detection based on impedance characteristicsBy studying the eddy current effect or magnetic effect caused by the non-ferromagnetic metal and ferromagnetic metal entering the alternating magnetic field,the change of the impedance parameter caused by the metal entering the coupled magnetic field is analyzed;the change rule of the impedance parameter caused by each variable is obtained;(2)Analysis and design of detection coilThe shape of the detection coil,the winding method,and the placement method were studied,and the coupling magnetic field distribution in various situations was analyzed,and the appropriate coil specifications were finally adopted;(3)Research on machine learning methodsCollect the amplitude and phase change data of various foreign objects,determine a reasonable parameter change threshold,and then analyze and study the data preprocessing method.A foreign object detection learning model based on machine learning is designed,and features such as feature extraction,model construction,and communication algorithms are analyzed and designed.The data is divided into sample data,the main model is constructed,and the training model enables it to quickly determine the type of foreign body and optimize the model.The serial communication algorithm implemented in the Python environment is compiled,real-time communication between the peripheral detection board and the Windows system on the PC side is realized,so that the PC side can quickly receive the foreign object alarm transmitted by the detection circuit and enter the machine learning model to judge the type of foreign object;...