Fault Diagnosis Method of Lithium-Ion Battery Leakage Based on Electrochemical Impedance Spectroscopy

Abstract

Electrolyte leakage may cause lithium-ion battery performance degradation, and even lead to short-circuit, resulting in serious safety accidents. In order to improve the safety of lithium-ion battery, it is necessary to detect electrolyte leakage in time. This paper presents a fault diagnosis method for electrolyte leakage of lithium-ion based on support vector machine (SVM) by electrochemical impedance spectroscopy (EIS) test. And the distribution of relaxation time (DRT) method is also employed to analyze the effect of leakage on the dynamic reaction process with full and half cells. In the proposed method, the EIS and DRT features are all used for the diagnosis to improve accuracy. Furthermore, the sensitivity of different features, the accuracy of battery classification under different cycles and state of charge were also analyzed to reduce diagnostic time by decreasing the test frequency range and SOC points and then validate the feasibility of the method.