A Study of incremental capacity analysis discrete wavelet transform-based feature extraction with stochastic analysis

The remaining-useful-life (RUL) prediction of lithium-ion batteries (LIBs) is crucial challenge for electric vehicles (EVs) and energy storage systems (ESSs). The health indicator (HI)-based RUL prediction have been investigated with experimental, model-based, and data-driven method. Most of the research have been focused on only one domain such as time or frequency domain. However, LIBs are complex system that operates in time (current and voltage) and frequency (electrochemical process; charge transfer, diffusion) domain. In this paper, incremental capacity analysis (ICA), which can analyze electrochemical properties, was performed to improve the interpretation of the constant current charging voltage curve in the time domain. In addition, discrete wavelet transform (DWT) with 7 mother wavelet functions were performed to extract features for the failure prevision.