Work

Incremental capacity feature selection for lithium-ion battery state of health estimation considering estimation capability and efficiency

Year: 2026

Type: article

Abstract: Accurate and transferable estimation of battery state of health is essential for the safety and reliability of electric vehicles and energy storage systems. However, many existing approaches rely on complete charging and discharging data and overlook how feature selection, robustness, and data requirements affect estimation performance. Here, we report a unified evaluation framework for predictive capability, transferable capability, and data efficiency for five features extracted from incremental capacity curves using partial charging data. We show that the voltage and magnitude of the first peak provide a better combination of accuracy, robustness across charge rates and temperatures, and minimal data needs. We demonstrate that these two features enable accurate estimation across two datasets. The results reveal that reliable health estimation can be achieved using only the portion of charging data corresponding to roughly less than 50% of the charge process, reducing data curation effort while maintaining high accuracy and practical transferability.

Source: Cell Reports Physical Science

DOI: 10.1016/j.xcrp.2025.103083

Authors: Lin Su, Shengyu Tao, Yingjie Chen, Changfu Zou, Xuan Zhang

Institutions: Tsinghua Shenzhen International Graduate School, Tsinghua University, Chalmers University of Technology, Peking University

Language: English

Topic: Advanced Battery Technologies Research

Subfield: Automotive Engineering

Field: Engineering

Domain: Physical Sciences

SDGs: Affordable and clean energy, Industry, innovation and infrastructure

Open Access status: gold

Estimated APC paid: $5,620