Remaining useful life prediction of lithium-ion batteries based on indirect feature and bidirectional long and short-term memory network optimized by beluga whale optimization

Lithium-ion batteries are widely used in new energy vehicles and other fields. The prediction of remaining useful life (RUL) is an essential and critical element of lithium-ion battery management systems. In this paper, we propose a framework for predicting the RUL of lithium-ion batteries using the beluga whale optimization (BWO) to optimize bidirectional long and short-term memory (BiLSTM) networks, aiming to improve the prediction accuracy of the RUL. Among them, the BWO is able to realize the automatic optimization of the network, which saves the manpower and time consumed in manual optimization while improving the accuracy effectively. To solve the problem that the capacity of lithium-ion batteries cannot be directly measured during use, indirect feature is used as a predictive feature. Firstly, several health indicators (HIs) are extracted during the battery usage. Afterwards, according to the Pearson's correlation coefficient between the HIs and the capacity, the HIs with the highest correlation are selected as the prediction feature, which ensures the reliability of indirect prediction. In this study, we conducted experiments using publicly available datasets and our own experimental datasets to verify the generalizability and robustness of the proposed method. The proposed method can realize the accurate prediction of the remaining service life of lithium-ion batteries online, which has practical application value.