A multi-model fusion method for state of health estimation of marine lithium-ion batteries

With the rapid development of electric ship technology, the state of health (SOH) estimation of marine lithium-ion batteries (LIBs) has become a key technology to ensure the safe and efficient operation of electric ships. However, currently, the research on predicting battery SOH under complex ship operating conditions is still very limited. In this paper, a multi-model fusion method is proposed to address the prediction of the battery SOH under complex ship operating conditions. Firstly, health indicators (HIs) that are easy to obtain during the battery rest state are selected based on the ship operating conditions. Secondly, the K-means clustering algorithm is used to adaptively divide the battery aging process into different aging intervals. Finally, multiple machine learning models are constructed, and by adopting the multi-model fusion method, the optimal SOH prediction model is selected for each aging interval, thereby improving the accuracy of the SOH prediction of LIBs. Additionally, a constant current discharge condition is set up to demonstrate the universality of the method.