Modeling and prediction of conducted EMI in on-board charging system of electric vehicle

Abstract

This paper focuses on the investigation of conducted electromagnetic interference (EMI) in a two-level onboard charging system for electric vehicles. It analyzes the mechanisms and coupling paths of conducted EMI in the system, identifying that EMI primarily originates from the switching actions of power components and propagates through transformers, passive components, and parasitic parameters of the circuit. Different models for various components are established using analytical methods, finite element numerical analysis methods, and measurement methods. Relevant parasitic parameters are extracted to construct a comprehensive system-level simulation and prediction model for conducted EMI. The accuracy of the simulation and prediction model is validated through EMI testing. To further predict the sources of conducted EMI in the system, an EMI node prediction method is proposed based on the simulation model. This method involves analyzing the frequency-domain simulation waveforms of EMI nodes within the charging system, identifying prominent nodes with significant EMI, and implementing relevant suppression measures for the components surrounding those nodes that contribute to EMI. The effectiveness of the proposed EMI node prediction method is verified through EMI suppression experiments.