BatteryMFormer: Multi-level Learning for Battery Degradation Trajectory Forecasting

arXiv:2605.27044 · 2026-05-28 공개 · arXiv · PDF

transformer trajectory-forecasting state-of-health battery-degradation soc-localization meta-pattern-memory aging-condition-aware dual-view-encoder

Abstract

Early battery degradation trajectory forecasting (BDTF), which predicts the full-life state-of-health trajectory from early operational data, is critical for battery optimization, manufacturing, and deployment. Battery degradation data exhibit two key characteristics. First, degradation data present a multi-level structure, including regularities shared within aging conditions and trajectory patterns shared across batteries. Second, degradation-related variations in voltage-current profiles are often localized to specific state-of-charge (SOC) intervals. Existing approaches often fail to explicitly model these characteristics. To bridge this gap, we propose BatteryMFormer, a multi-level Transformer for early BDTF. BatteryMFormer integrates (1) an aging-condition-aware decoder that injects aging-condition priors via aging-condition-informed queries and aging-condition-aware attention, (2) a meta degradation pattern memory that learns and retrieves trajectory prototypes to guide long-horizon forecasting, and (3) a dual-view encoder that jointly captures temporal dynamics and SOC-localized variations from voltage and current time series. Extensive experiments on four battery domains show that BatteryMFormer consistently outperforms state-of-the-art baselines, marking a significant step toward reliable BDTF. Our code is available at https://github.com/Ruifeng-Tan/BatteryMFormer.

한국어 요약

📋 한 줄 요약

**[Battery SOH Forecasting / Transformer]** BatteryMFormer가 aging-condition-aware decoder·meta degradation pattern memory·dual-view encoder로 multi-level 구조 명시 모델링, 4 배터리 도메인에서 SOTA early BDTF.

🎯 핵심 기여도

💡 핵심 아이디어

Early BDTF의 정확도는 배터리 데이터의 multi-level 구조(aging condition·trajectory 공유)와 SOC-localized 변동성을 명시적으로 모델링하는 Transformer 아키텍처에서 비롯되며, condition-aware query·meta pattern memory·dual-view encoding의 통합이 단일 표현으로는 잡지 못하는 구조를 포착한다.

🔬 기술적 접근법

📊 주요 결과

💭 의의 및 한계

**의의**: 배터리 degradation 데이터 특성에 맞춘 첫 multi-level Transformer 설계, aging-condition·SOC-localized 모델링의 명시화, 실용 응용(최적화·제조·배포) 가치 큼. **한계**: 4 도메인 중심으로 매우 다른 화학·셀 화학의 일반화는 추가 검증, early operational data의 길이·품질 의존, dual-view encoder의 계산 비용.

🚀 실용적 활용