High-reliability multi-fault diagnosis of lithium-ion batteries based on low-redundancy cross-measurement and affine transformation

Electrical faults pose significant risks to the safety of battery packs. The cross-voltage measurement circuit (CVMC) offers a common solution for diagnosing multiple types of electrical faults. However, balancing diagnosis reliability with sensor redundancy in CVMC remains a challenging problem. Motivated by this, we propose a low-sensor redundancy CVMC designed for the reliable diagnosis of the common electrical faults, including the internal short circuit, connection faults, and sensor faults. Specifically, each sensor in the proposed CVMC sequentially monitors three neighboring components (two cells and one connection plate, or one cell and two connection plates), ensuring the overlapping measurements for each cell and connection plate. Moreover, affine transformation with multiple independent elements is exploited to delicately characterize faults, greatly enhancing the pointing to specific faults. By integrating both methods, fault types and locations can be accurately distinguished and determined. Experimental results show the effectiveness and reliability of the proposed multi-fault diagnosis method.