基于电池电压极值特征融合的采样故障诊断

打开文本图片集

主题词：新能源汽车安全动力电池电压采样故障特征融合中图分类号：U463.633；TP277 文献标志码：A DOI：10.19620/j.cnki.1000-3703.20250342

Sampling Fault Diagnosis Based on Features Fusion Derived from BatteryVoltageExtremeValue

Ren Yonghuan1，Cheng Zhen'，Zheng Binbin¹²，Sun Weijial2 （1.XiamenKingLong UnitedAutomotiveIndustryCo.，Ltd.，Xiamen;2.KeyLaboratoryofBusSafetyandPower saving Technology Enterprises ofFujian Province，Xiamen ）

【Abstract】To address the high false positiverate and weak anti-interference capabilitiesof existing voltage Sampling Fault （SF）diagnosticmethods，thispaper proposesanovel diagnosticalgorithm thatutilizes modelresidual meanvalueand residual standarddeviation derivedfromvoltage extreme valueas feature values combined with covariance coeffcients to identifybattryvoltage SF.Thismethodachieves99.3%recognized precisionand0.91Intersectionover Union，ffectively resolving theproblemthatthebattery managementsystemscannotdistinguish SFfrombattrycellisses.Itcontributes to intelligent identification of battery safety risk levels and enhances the safety of electric vehicles.

KeyWords：Safety of electric vehicles，Powerbattery，Voltage samplingfault，Featuresfusion

【引用格式】任永欢，程震，郑彬彬，等.基于电池电压极值特征融合的采样故障诊断[J].汽车技术，2025（9）：10-16. RENY H，CHENG Z，ZHENG BB，etal.Sampling Fault Diagnosis BasedonFeatures Fusion Derived from Batery Voltage ExtremeValue[J]. Automobile Technology，2025（9）：10-16.

1前言

电池管理系统（BatteryManagementSystem，BMS）电压类报警依托于电池电压的监控，当发生电压采样故障（SamplingFault，SF）时，电压采样值因异常波动超出报警阈值亦会引发过欠压报警。（剩余8956字）