Internal short circuit detection and evaluation in battery packs based on transformation matrix and an improved state-space model

Abstract

Internal short circuit (ISC) is a serious safety hazard for lithium-ion battery packs. How to comprehensively detect and evaluate ISC in battery packs remains a challenging problem. Motivated by this, this paper proposes an ISC detection method based on the transformation matrix and an ISC resistance calculation method based on an improved state-space model. Specifically, benefiting from the descriptive ability of the transformation matrix for curve changes, the shear element in the transformation matrix is used to capture the skewed and downward voltage pattern of the ISC cell in the battery pack. An online detection flow is designed based on the opposite variation relationship between the shear elements calculated from adjacent cells. Moreover, an improved state-space model is developed to directly estimate the ISC current , which ensures that the weight of the ISC current in the model is unrestricted. A dual extended Kalman filter with separated time scales is deployed to obtain accurate state estimations. Experimental results validate the effectiveness and advantages of the ISC detection method, as well as the high accuracy of the ISC resistance calculation method.