Internal short circuit (ISC) is the main cause of thermal runaway (TR) in lithium-ion batteries, and early detection of ISC is crucial to improve battery safety. This paper introduces a method for detecting ISC and classifying fault severity by analyzing the variations in voltage and surface temperature during battery operation. Firstly, the electrothermal coupling model (ETCM) of the battery was constructed, and the Pearson correlation coefficient (PCC) was used to identify the ISC of the battery through the real-time collected voltage and temperature sensor data When the battery was working. Then the battery model with ISC is updated by using equivalent ISC resistance. The battery status estimation is achieved by integrating the extended Kalman filter (EFK) and sliding mode observer (SMO). Finally, the fault grade is classified according to the state difference between the battery with ISC and normal battery. The batteries with different degrees of ISC were installed in battery packs and verified by the StarSim hardware-in-the-loop (HIL) experiment. The findings suggest that the proposed approach facilitates rapid identification of ISC in the battery pack and enables categorization of fault severity based on the state estimation outcomes derived from analyzing the battery with ISC.
Read full abstract