Foundation Models Knowledge Distillation For Battery Capacity Degradation Forecast

TL;DR

This paper develops a foundation model for lithium-ion battery capacity degradation forecasting, demonstrating improved accuracy and generalization across scales and conditions, and introduces knowledge distillation for efficient deployment.

Contribution

It proposes a degradation-aware fine-tuning strategy for large pre-trained time-series models and applies knowledge distillation to create compact, efficient models for battery health prediction.

Findings

Battery-Timer outperforms specialized models in capacity prediction.

Knowledge distillation reduces computational costs while maintaining accuracy.

Model generalizes well across different battery scales and operating conditions.

Abstract

Accurate forecasting of lithium-ion battery capacity degradation is critical for reliable and safe operation, yet remains challenging under distribution shifts across scales and operating regimes. Here we investigate a time-series foundation model, that is, a large pre-trained time-series model for capacity degradation forecasting, and propose a degradation-aware fine-tuning strategy that aligns the model to capacity trajectories while retaining broadly transferable temporal structure. We instantiate this approach by fine-tuning the Timer model on 220,153 cycles of open-source charge-discharge records to obtain Battery-Timer. Using our released CycleLife-SJTUIE dataset, a real-world industrial collection from an energy-storage station with long-horizon cycling, we evaluate capacity generalization from small cells to large-scale storage systems and across varying operating conditions. Battery-Timer consistently outperforms specialized expert models. To address deployment cost, we further introduce knowledge distillation, a teacher-student transfer that compresses the foundation model's behavior into compact expert models. Distillation across several state-of-the-art time-series experts improves multi-condition capacity generalization while substantially reducing computational overhead, indicating a practical path to deployable cross-scale degradation forecasting by combining a foundation model with targeted distillation.