EV Diagnostics and Anomaly Detection System for Internal Faults in Electric Vehicles

“EV Diagnostics and Anomaly Detection System for Internal Faults in Electric Vehicles”

Vishal Chaudhari1 Azim Bagwan2 Sanjana Choudhary3 Rasika Gaikwad4

under the guidance of Prof. Ashwini Kamble

1-3Computer Engineering, K J College of Engineering and Management Research, Pune, India

ABSTRACT

The increasing focus on the safety of lithium-ion batteries (LIBs) in electric vehicles (EVs) necessitates early detection of soft short circuits (SCs) to prevent severe faults such as fire or thermal runaway. This paper proposes an on-board soft SC fault diagnosis method using the extended Kalman filter (EKF). The EKF adjusts a gain matrix based on real-time measured voltages to estimate the state of charge (SOC) of the faulty cell. The SOC difference is then utilized for soft SC fault detection, with identified soft SC resistance values indicating the severity of the fault. Experimental validation on a series-connected battery pack confirms the method's effectiveness in promptly detecting soft SC faults and accurately estimating their resistance.

While lithium-ion batteries offer advantages such as high energy density and quick charging, concerns about their thermal stability hinder their widespread use due to potential fire and explosion risks. This manuscript comprehensively reviews the thermal runaway phenomenon and fire dynamics in both single LIB cells and multi-cell battery packs. It discusses potential fire prevention measures, emphasizing the challenges associated with ensuring the safety of LIB applications in electric vehicles and energy storage systems. Additionally, the paper provides an overview of fault detection methods for critical EV components, including Permanent Magnet Synchronous Motors (PMSMs) and lithium-ion battery packs, stressing the importance of accuracy, speed, sensitivity, and cost-effectiveness in fault detection approaches, with a focus on the latest research developments.