State of Health Estimation and Remaining Useful Life Estimation for Li-ion Batteries Based on a Hybrid Kernel Function Relevance Vector Machine

Abstract

Accurate estimations of the state of health (SOH) and the remaining useful life (RUL) of lithium batteries are important indicators that ensure the safe and stable operation of a battery system. However, these two health indicators are difficult to estimate during online operation. This paper proposes a joint estimation method for SOH and RUL based on the hybrid-kernel RVM (H-RVM) method. The method extracts the segment data features of the charging voltage, current and temperature, online,by analysing the battery incremental capacity (IC) curve and obtains the indirect health factor (IHF) by reducing the dimension through a principal component analysis (PCA). Then, an ageing model of a lithium battery is established by the RVM algorithm. On this basis, another RVM is used to perform a multistep prediction of the IHF, combining the prediction results with the battery ageing model and comparing the failure thresholds to attain a RUL estimation. Finally, three groups of battery data, under different ageing conditions, are used for verification. The results show that the method proposed in this paper has high accuracy and stability.