Multi-dimension Fault Diagnosis of Battery System in Electric Vehicles Based on Real-world Thermal Runaway Vehicle Data

Battery system fault diagnosis and thermal runaway warning are critical to ensure the safe operation of electric vehicles. In this paper, the multi-dimensional fault diagnosis method is used to quantitatively analyze the whole-life-cycle data of real-world thermal runaway vehicle. Firstly, the battery voltage range diagnosis, outlier unit identification, voltage jump diagnosis and temperature range diagnosis are carried out to mine the potential fault of the power battery. Then the cell voltage and probe temperature diagnosis of the charge and discharge segment before thermal runaway are performed to accurately locate thermal runaway potential cells and provide early warning of thermal runaway. The conclusion of the fault diagnosis is verified by battery disassembling analysis. The results show that the potential failure of the cell No. 69 is obvious. During the last charging segment before thermal runaway, a large voltage drop occurred to the cell No. 69, and the temperature of temperature probe near it in the same module rose too fast. This method can accurately locate and warn potential thermal runaway cells one day before thermal runaway.