Abstract Accurate remaining useful life (RUL) prediction for lithium-ion batteries (LIBs) is a key technology for ensuring the safety of energy storage systems such as electric vehicles. It can not only prevent safety hazards caused by battery aging or failure but also provide a scientific basis for battery health management. However, several prominent problems exist with current RUL prediction models. First, LIBs are susceptible to various uncertain factors, which makes it difficult to guarantee the accuracy of the models. Second, the modeling process and prediction results have insufficient interpretability, and fail to meet the transparency requirements for models in the field of LIB safety. To address these challenges, a LIB RUL prediction method is proposed based on the belief rule base (BRB) with a balance between interpretability and accuracy (BRB-IA). First, the key features significantly related to the health status of LIBs are selected from multidimensional health indicators (HIs) through correlation analysis methods. Second, six interpretability guidelines, three optimization strategies, and three constraint methods are introduced to maintain both interpretability and accuracy in the BRB-IA model. Moreover, an evaluation criterion for the interpretability of RUL prediction models is designed. Finally, the effectiveness and superiority of the proposed method were analyzed and validated through RUL prediction experiments using public LIB datasets from both NASA and Xi’an Jiaotong University.
Yang et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: