State-of-Health Prediction for EV Lithium-Ion Batteries via DLinear and Robust Explainable Feature Selection

TL;DR

This paper presents an explainable, data-driven framework for predicting the state-of-health of EV lithium-ion batteries, combining robust feature selection with a DLinear model to improve accuracy and efficiency in battery management.

Contribution

It introduces a novel SOH prediction method that integrates SHAP-based feature selection with DLinear, effectively capturing cell-to-cell variability and outperforming traditional models.

Findings

DLinear achieves higher accuracy than LSTM and Transformer models.

Selected features are consistent across multiple cells, capturing cell-to-cell variability.

The framework is computationally efficient and suitable for real-time EV battery management.

Abstract

Accurate prediction of the state-of-health (SOH) of lithium-ion batteries is essential for ensuring the safety, reliability, and efficient operation of electric vehicles (EVs). Battery packs in EVs experience nonuniform degradation due to cell-to-cell variability (CtCV), posing a major challenge for real-time battery management. In this work, we propose an explainable, data-driven SOH prediction framework tailored for EV battery management systems (BMS). The approach combines robust feature engineering with a DLinear. Using NASA's battery aging dataset, we extract twenty meaningful features from voltage, current, temperature, and time profiles, and select key features using Pearson correlation and Shapley additive explanations (SHAP). The SHAP-based selection yields consistent feature importance across multiple cells, effectively capturing CtCV. The DLinear algorithm outperforms long short-term memory (LSTM) and Transformer architectures in prediction accuracy, while requiring fewer training cycles and lower computational cost. This work offers a scalable and interpretable framework for SOH forecasting, enabling practical implementation in EV BMS and promoting safer, more efficient electric mobility.