An Applicable Minor Short-Circuit Fault Diagnosis Method for Automotive Lithium-Ion Batteries Based on Extremum Sample Entropy

Maintaining the safety of lithium-ion battery modules is the priority in promoting the application of electric vehicles (EVs). In practical EV applications, only the total voltage of the battery module and the maximum/minimum cell voltages are available. Under this circumstance, most existing methods are unable to diagnose the faults of EV battery modules. Therefore, an applicable minor short-circuit fault diagnosis method for automotive lithium-ion batteries based on extremum sample entropy (ESE) is proposed to solve the above issues. Specifically, the extremum sequences are first extracted from the original voltage data. Then, the sample entropy of the sum/difference sequence is applied to diagnose minor cell internal short-circuit (C-ISC) faults and module external short-circuit (M-ESC) faults. Eventually, the mean extreme difference model is established to quantitatively evaluate the internal short-circuit (ISC) resistances. The experimental results reveal that the proposed ESE algorithm can successfully diagnose the different degrees of minor C-ISC and M-ESC faults. Moreover, the average value of the estimated ISC resistance is 71 m omega, whose estimated error is 5.6%. More importantly, the proposed ESE approach requires only 1 s to detect 100 battery cells, which increases the calculation speed by 30 times compared with the traditional method.