Battery State of Health Estimation Using LLM Framework

TL;DR

The paper presents a transformer-based multimodal LLM framework that combines cycle-based and instantaneous discharge data with Differential Voltage Analysis (DVA) to estimate SoH and predict RUL for lithium titanate (LTO) batteries. It demonstrates high predictive accuracy, achieving MAE as low as $0.87\%$ in the abstract and $0.81\%$ in comparative metrics, while emphasizing anomaly detection for early degradation signals and predictive maintenance. Although the approach delivers strong accuracy, it incurs higher inference latency, motivating a hybrid edge-cloud deployment to balance real-time monitoring with more compute-intensive analysis. The work advances battery health monitoring by integrating DVA-informed features into an LLM pipeline, showing potential for real-time EV battery management and proactive maintenance strategies.

Abstract

Battery health monitoring is critical for the efficient and reliable operation of electric vehicles (EVs). This study introduces a transformer-based framework for estimating the State of Health (SoH) and predicting the Remaining Useful Life (RUL) of lithium titanate (LTO) battery cells by utilizing both cycle-based and instantaneous discharge data. Testing on eight LTO cells under various cycling conditions over 500 cycles, we demonstrate the impact of charge durations on energy storage trends and apply Differential Voltage Analysis (DVA) to monitor capacity changes (dQ/dV) across voltage ranges. Our LLM model achieves superior performance, with a Mean Absolute Error (MAE) as low as 0.87\% and varied latency metrics that support efficient processing, demonstrating its strong potential for real-time integration into EVs. The framework effectively identifies early signs of degradation through anomaly detection in high-resolution data, facilitating predictive maintenance to prevent sudden battery failures and enhance energy efficiency.

Figures (4)

• Figure 1: Pipeline for State of Charge (SoC) Prediction Using Multi-Input Regression with Schema-Based Data Processing and Model Evaluation
• Figure 2: End-to-End Pipeline for State of Charge (SoC) Prediction: Data Handling, Model Training & Validation, and Evaluation
• Figure 3: Pipeline for Multi-Input Regression Model in Battery State Estimation: Data Handling, Training, and Evaluation
• Figure 4: Pipeline for Multi-Input Regression Model Training and SoC Prediction with Comprehensive Data Handling and Evaluation