Research on Lithium-Ion Battery State of Health Prediction Based on XGBoost–ARIMA Joint Optimization
Due to the complex electrochemical reactions within lithium-ion batteries and the uncertainties with respect to external environmental factors, accurately assessing their State of Health (SOH) remains a significant challenge. To improve the precision of SOH estimation, we propose an intelligent esti...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Batteries |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-0105/11/6/207 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Due to the complex electrochemical reactions within lithium-ion batteries and the uncertainties with respect to external environmental factors, accurately assessing their State of Health (SOH) remains a significant challenge. To improve the precision of SOH estimation, we propose an intelligent estimation approach that integrates data visualization and advanced machine learning techniques. Initially, the battery data are visualized using matplotlib to extract key features such as temperature difference, voltage difference, and average voltage. Subsequently, an XGBoost-based model is constructed to perform the initial SOH estimation. To further enhance the estimation accuracy, we introduce the Autoregressive Integrated Moving Average Model (ARIMA) model for post-estimation correction, effectively refining the preliminary results. Experimental results demonstrate that the proposed XGBoost–ARIMA model outperforms traditional algorithms, including Linear Regression (LR), Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN), not only in estimation accuracy but also in generalization capability, showing significant improvements over five other regression models. |
|---|---|
| ISSN: | 2313-0105 |