Remaining Useful Life Prediction Based on Nonlinear Wiener Process and Time-Varying Correlation Model for Lithium-Ion Batteries

Lithium-Ion batteries are widely used in a variety of critical applications due to their excellent energy density and low self-discharge rate. Ensuring their reliable operation is essential for both economic performance and operational safety. Among the various tasks in battery health management, accurately predicting the Remaining Useful Life (RUL) of lithium-ion batteries is one of the most critical. However, current RUL prediction methods face significant challenges due to high nonlinearity, data dispersion, and coupled degradation among monitored parameters. To address these issues, this study proposes an RUL prediction method based on a nonlinear Wiener process and time-varying correlation model for lithiumion batteries. An improved Markov Chain Monte Carlo (MCMC) method is introduced to achieve more accurate parameter estimation of the nonlinear Wiener process. And an adaptive time-varying Copula modeling method is employed to capture the dynamic correlations among degradation parameters. Experimental results demonstrate that the proposed method effectively estimates the nonlinear Wiener process parameters for two key degradation parameters: discharge capacity and internal resistance. Furthermore, modeling their time-varying correlation structure improves the RUL prediction accuracy.