Abstract:Battery health prognostic plays a crucial role in the safe operation of electrical devices. Battery degradation modeling is a crucial step for battery health prognostics. Currently, since most existing battery prognostic models only consider entire long-term degradation trends during modeling, battery degradation modeling is still not precise and lacks short-term deviations to sufficiently reflect strong complexity and volatility of degradation processes. To solve this problem, a new battery prognostic method is proposed by simultaneously considering long-term degradation tendency and short-term volatility. Because Wiener process and Ornstein–Uhlenbeck process have ability to respectively model long-term degradation tendency and short-term volatility, a short-long-term degradation model is accordingly proposed by linearly fusing two stochastic processes. Additionally, the proposed model can characterize modeling uncertainty and observation uncertainty. Besides, this article proves that the proposed model follows a log-normal distribution, which provides solid foundations for state-space modeling. Here, the purpose of formulating the proposed model to a state-space model with multiple cross-correlation hidden state variables is to better describe individual battery degradation using online battery capacity data. Health prognostics can be easily realized by extrapolating the updated model to a failure threshold. Results show that our proposed method performs well than contrast methods for battery health prognostics.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper aims to address the accuracy issue in battery health prediction. Specifically, most existing battery prediction models only consider the overall long-term degradation trend while ignoring short-term fluctuations, leading to less precise battery degradation modeling that fails to fully reflect the complexity and volatility of the degradation process. To solve this problem, the paper proposes a new battery prediction method that improves the accuracy of battery health prediction by considering both long-term degradation trends and short-term fluctuations simultaneously. ### Main Contributions 1. **Integration of Long-term Degradation Trends and Short-term Fluctuations**: - The proposed new short-long-term degradation model is a linear combination of the Wiener Process (WP) and the Ornstein–Uhlenbeck Process (OUP), which can flexibly model both long-term degradation trends and short-term fluctuations simultaneously, achieving precise battery degradation modeling. - Compared to AI-based and model-based methods, this model better describes modeling uncertainty and observation uncertainty. 2. **Theoretical and Mathematical Support**: - It is proven that the proposed model follows a log-normal distribution, and the analytical expressions of its key parameters (mean and covariance) are derived, providing a solid theoretical foundation for the construction of the state-space model. 3. **Adaptive Update for Individual Battery Degradation**: - A multivariate cross-correlated hidden state-space model is designed, and combined with Kalman filtering and online capacity degradation data, it adaptively updates multiple correlated state variables to better predict the degradation of individual batteries. - The model's hyperparameters are automatically adjusted through maximum likelihood estimation, eliminating the need for expert knowledge and manual settings. ### Experimental Validation The paper conducts experimental validation using two different battery degradation datasets, namely experimental dataset A and public dataset B. The experimental results show that the proposed method outperforms two well-known existing prediction methods, including the Wiener Process and the Boxcox-based method, in terms of battery health prediction. ### Conclusion The new short-long-term degradation model proposed in the paper significantly improves the accuracy of battery health prediction by considering both long-term degradation trends and short-term fluctuations. The model performs excellently both theoretically and in practical applications, providing strong support for battery health management and prediction.

New Short-Long-Term Degradation Model for Precise Battery Health Prognostics

A Physics-Informed Hybrid Data-Driven Approach with Generative Electrode-Level Features for Lithium-Ion Battery Health Prognostics

Enhanced Gaussian Process Dynamical Modeling for Battery Health Status Forecasting

Data-Driven Battery Health Prognosis Using Adaptive Brownian Motion Model

Accelerated Stress Factors based Nonlinear Wiener Process Model for Lithium-ion Battery Prognostics

A Neural-Driven Stochastic Degradation Model for State-of-health Estimation of Lithium-Ion Battery

A Hierarchical Model for Lithium-Ion Battery Degradation Prediction

Joint nonlinear-drift-driven Wiener process-Markov chain degradation switching model for adaptive online predicting lithium-ion battery remaining useful life

Degradation Modeling Considering the Dependency of Rate and Volatility for Real-Time Prognostics with Error Correction

Battery Health Prognosis Using Brownian Motion Modeling and Particle Filtering

A novel based-performance degradation Wiener process model for real-time reliability evaluation of lithium-ion battery

Stochastic modeling for tracking and prediction of gradual and transient battery performance degradation

Li-Ion Battery State of Health Estimation and Remaining Useful Life Prediction Through a Model-Data-Fusion Method

A rest-time-based prognostic model for remaining useful life prediction of lithium-ion battery

Degradation Trajectory Prognostics for Lithium-Ion Batteries under Varied Operating Conditions with a Hybrid-Driven Method.

Li-ion Battery Prognostic and Health Management Through an Indirect Hybrid Model

A State-Space-Based Prognostic Model for Hidden and Age-Dependent Nonlinear Degradation Process

Voltage-temperature health feature extraction to improve prognostics and health management of lithium-ion batteries

Remaining Useful Life Prediction for Degradation with Recovery Phenomenon Based on Uncertain Process

Lithium-Ion Battery Prognostics with Hybrid Gaussian Process Function Regression

Hybrid physics-based and data-driven modeling with calibrated uncertainty for lithium-ion battery degradation diagnosis and prognosis