Big-data Analysis and Knowledge Discovery of Battery Fault in Numerous Real-world Electric Vehicles

Abstract

Battery safety and aging have been considered as most import issues restricting the further deployment and development of electric vehicles in real-world applications. Due to differences in technology, materials, groups and physical and chemical reactions inside the battery, lithium-ion concentration gradient will be formed inside the battery, which is directly reflected in the inconsistency of external parameters.To study the correlation between battery faults and external parameters, in this study, alllife-cycle big-data of 41 electric vehicles are analysed by data mining. Firstly, the characteristic parameters representing voltage consistency are studied, and then the selected related parameters are statistically analysed in each dimension to explore the factors affecting the safety and reliability of batteries. The interesting knowledge discovered by this study can provide follow-up support for battery safety.