电动汽车充电接口可靠性与故障预测模型

打开文本图片集

中图分类号：U469.72 文献标识码：A 文章编号：1003-8639（2026）01-0023-03

Reliability and Fault Prediction Model of Electric Vehicle Charging Interface

Liu Tao

（Shandong Traffic Technician College，Linyi 276OoO，China）

【Abstract】This study focuses on enhancing the electrical contact reliability of electric vehicle charging interfacesandconstructing afaultpredictionmodel.Byanalyzingthefailure mechanisms of electricalcontacts，reliability improvement schemes such as modular wiring harness design and anti-corosion coating application are proposed. A dynamic monitoring method for contact resistance is established based on IEC standards.A GCN-LSTM deep learning model isconstructed tointegratecurrent/voltage time-seriesdata withuserbehaviorfeatures，enabling earlyfault warning. Experimental validation demonstrates that the model achieves an accuracy of 88.12% and an F1 Score of O.844 in charging pile fault diagnosis，outperforming traditional models.This research provides theoretical supportand technical pathways forinteligentoperationand maintenanceofcharging facilities，whichisof great significanceforensuring the healthy development of the new energy vehicle industry.

【KeyWords】charging interface;electrical contactreliability；fault prediction model;multimodal fusion： GCN-LSTM

0 引言

随着全球新能源汽车保有量突破2亿辆，充电接口作为能量传输的核心枢纽，其电气接触可靠性直接影响用户安全与电网稳定。（剩余4470字）