Minor Faults Diagnosis for Under-Sampled Lithium-Ion Batteries Based on Static-Dynamic Compensation

Abstract

With the rapid proliferation of electric vehicles, the safety concerns related to lithium-ion batteries are gaining more and more attention. Fault diagnosis is a key approach to reducing the risk of battery failure. However, existing battery management systems (BMS) apply under-sampled voltage signal acquisition, which leads to misdiagnosis and omission of faults. To address this issue, a minor fault early diagnosis method based on static-dynamic compensation voltage data is proposed. First, the voltages of the series-connected cells are asynchronously collected. Then, the collected voltage sequences from various modules are mapped to the voltage sequence of the target battery using the static-dynamic compensating method, which can obtain a new sequence with a significantly higher equivalent sampling frequency. Finally, the sample entropy method is employed to detect minor faults based on the new sequence after compensation. Experimental results reveal that the presented method can increase the sampling frequency by about 8 times. The proposed method can successfully detect minor short circuits and poor connection faults in the battery under different ambient temperatures.