Safety Risk Identification of Lithium-ion Battery Based on Kolmogorov Complexity

Battery is the key component and main trouble source of an electric vehicle (EV). With the rapid growth of market share, thermal runaway caused by malfunction of batteries have been frequently reported, so fault diagnosis is critical to ensure safety and to improve performance. Unfortunately, most of the existing fault diagnosis methods only focus on the identification of voltage anomalies on single cell level, ignoring the characteristics on macro system level. Consequently, without obvious abnormality in voltage, faults of certain types can hardly be caught. This paper proposes a novel fault diagnosis method based on Kolmogorov complexity, which can quantitatively describe the degree of confusion over battery pack level to identify potential risk. The proposed method is verified by real EVs operation data collected through the National Monitoring and Management Center for New Energy Vehicles, where clear correlation between the increased level of Kolmogorov complexity and thermal runaway is observed. As a simple conclusion, the proposed method can be an important supplement to traditional fault diagnosis methods.