An Online Adaptive Internal Short Circuit Detection Method of Lithium-Ion Battery

Abstract

Internal short circuit (ISC) is a critical cause for the dangerous thermal runaway of lithium-ion battery (LIB); thus, the accurate early-stage detection of the ISC failure is critical to improving the safety of electric vehicles. In this paper, a model-based and self-diagnostic method for online ISC detection of LIB is proposed using the measured load current and terminal voltage. An equivalent circuit model is built to describe the characteristics of ISC cell. A discrete-time regression model is formulated for the faulty cell model through the system transfer function, based on which the electrical model parameters are adapted online to keep the model accurate. Furthermore, an online ISC detection method is exploited by incorporating an extended Kalman filter-based state of charge estimator, an abnormal charge depletion-based ISC current estimator, and an ISC resistance estimator based on the recursive least squares method with variant forgetting factor. The proposed method shows a self-diagnostic merit relying on the single-cell measurements, which makes it free from the extra uncertainty caused by other cells in the system. Experimental results suggest that the online parameterized model can accurately predict the voltage dynamics of LIB. The proposed diagnostic method can accurately identify the ISC resistance online, thereby contributing to the early-stage detection of ISC fault in the LIB.