Evaluating fault detection strategies for lithium-ion batteries in electric vehicles

Abstract

Electric Vehicles (EVs) are a rapidly growing segment in India’s automotive sector, with an expected 70% growth by 2030. Lithium-ion (Li-ion) rechargeable batteries are favoured because of their high efficiency in power and energy delivery, along with fast charging, long lifespan, low self-discharge, and environmental friendliness. However, as a crucial subsystem in EVs, batteries are susceptible to faults arising from various factors. Li-ion battery faults can be categorized as internal or external. Internal faults stem from over-charging, over-discharging, overheating, acceleration and degradation processes, short circuits, and thermal runaway. External faults are caused by sensor malfunctions, cooling system failures, and cell connection problems. A Battery Management System (BMS) plays an essential role in regulating battery operation, monitoring its health status, and implementing fault diagnostic techniques. Fault diagnostic algorithms running on the BMS enable early or post-fault detection and control measures to minimize the consequences of faults, thereby ensuring battery safety and reliability. This paper reviews various internal and external battery fault diagnosis methods. In addition to battery fault detection, this work conducts a comparative analysis of optimization techniques for fault diagnosis, including Fuzzy Clustering, Long Short-Term Memory, Support Vector Machines, and Particle Swarm Optimization.