Research on short-circuit fault-diagnosis strategy of lithium-ion battery in an energy-storage system based on voltage cosine similarity

In recent years, accidents such as spontaneous combustion and explosion have frequently occurred in the field of electrochemical energy storage, and thermal runaway caused by short-circuit faults in lithium-ion (Li-ion) batteries is one of the main reasons. This study investigated the internal short circuit (ISC) fault diagnosis method for Li-ion (LiFePO4) batteries in energy storage devices. A short-circuit fault diagnosis method for battery module components based on voltage cosine similarity is proposed based on the characteristics extracted from the ISC fault battery. In this method, the voltage and current of the battery were used to derive a two-dimensional feature vector, and a gain multiple with excitation information was introduced to perform secondary processing on the eigenvectors and achieve effective separation of the fault-signal features. The experimental results show that the proposed method can achieve a second-level rapid diagnosis of short-circuit faults in batteries when the simulated ISC resistance is below 5 Ω, and the excellent fault-detection capabilities for modules with inconsistent states of charge and the performance of the proposed method under dynamic conditions are verified.