A Basic Study on the Diagnosis Method of Internal Short Circuit based on Feature Extraction and Separation Technique

Abstract

Lithium-ion batteries are widely used in energy storage systems (ESS) and electric vehicles due to superior performance. However, most safety problems in the use process of the lithium-ion battery are attributed to internal short circuits, there are a lot of factors leading to the internal short circuit. Short circuit in the lithium-ion battery affects the electrochemical reaction and then generating a larger short circuit current. Besides, the thermal runaway is the result of a failure of the separator inside Li-ion batteries, small discharge pulses of a few mV were detected at the time of separator damages. Therefore, the fault diagnosis by internal short circuit of lithium batteries has attracted more attention to prevent the fire accidents. This paper aimed to suggest a diagnostic model for separator of LFP (Lithium iron phosphate) pouch cell as a basic research work using continuous pulse analysis method such as feature extraction and separation technique of pulse source for timely diagnose the short circuit fault. In other words, this paper provided a method that can identify the separator failure before thermal runaway in the aspect of a potential explosion and fire prevention. The internal short circuit signal for diagnosis of LFP pouch cell was investigated by forming artificial defects at the separator part compared with normal battery. To investigate the short circuit signal, artificial defect specimens are prepared as a pouch type, pulse signal was analyzed in terms of pulse waveshape (Skewness) and frequency (Fast Fourier Transform, FFT) compared with normal battery and artificial defect pouch cell. To calculate the continuous acquisition pulse signal, python code was developed. As a result, it is found that the pulse signal of artificial defect specimen shown high frequency and low skewness characteristics. Whereas, in the case of a normal battery, it showed characteristics of relatively low frequency and high or middle skewness. From our research work, the results revealed that the pulse shape and feature contain useful information which can help to understand degradation behavior of the LFP pouch cells. This result is expected to be applicable various batteries with presence a defect at the separator of battery