Domain-adaptive state of health prediction of vehicle batteries powered by deep learning

Minghe Li; Zicheng Fei; Luoxiao Yang; Zijun Zhang; Kwok-Leung Tsui

doi:10.1016/j.xcrp.2025.102550

Domain-adaptive state of health prediction of vehicle batteries powered by deep learning

Minghe Li, Zicheng Fei, Luoxiao Yang, Zijun Zhang^*, Kwok-Leung Tsui

^*Corresponding author for this work

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

Abstract

State of health (SOH) estimation of battery packs in electric vehicles (EVs) is essential for transportation electrification safety and reliability. The noise and complexity of EV battery pack data hinder the effectiveness of various data-driven SOH estimation methods using lab data. To address these challenges and achieve more effective data-driven EV battery pack SOH predictions, this study develops a comprehensive deep-learning-based SOH modeling framework for EV batteries. The framework begins with a two-stage mode decomposition (TSMD) method designed to effectively identify neat SOH degradation patterns better representing noisy field data. Next, an endogenous and exogenous multibranch network structure with a hierarchically fused attention mechanism (EEMB-HiFA) is developed for real-time prediction of EV battery pack SOH. Computational experiments leveraging datasets from seven EVs are conducted to validate the accuracy and adaptiveness of the proposed EEMB-HiFA. The results show that the EEMB-HiFA can achieve a 96.49% improvement in accuracy compared to strong benchmarks considered. © 2025 The Author(s)

Original language	English
Article number	102550
Journal	Cell Reports Physical Science
Online published	17 Apr 2025
DOIs	https://doi.org/10.1016/j.xcrp.2025.102550
Publication status	Online published - 17 Apr 2025

Funding

This work was supported in part by the Guangdong Province Technological Project under grant 2023A0505030014, in part by the Shenzhen-Hong Kong-Macau Science and Technology Category C Project under grant SGDX20220530111205037, in part by the Hong Kong RGC General Research Fund Project under grant 11213124, in part by the Hong Kong RGC Collaborative Research Fund Project under grant C1049-24GF, in part by the Hong Kong ITC Innovation and Technology Fund Project under grant ITS/034/22MS, in part by the Guangdong Provincial Basic and Applied Basic Research-Offshore Wind Power Joint Fund Project under grant 2022A1515240066, and in part by the InnoHK Initiative, The Government of the HKSAR, and the Laboratory for AI-Powered Financial Technologies. Open Access made possible with partial support from the Open Access Publishing Fund of the City University of Hong Kong.

Research Keywords

data feature fusion
data-driven model
deep learning
domain adaptation
electric vehicles
lithium-ion battery
signal denoising
state of health

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access Output

10.1016/j.xcrp.2025.102550

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{5b4cca2fb33547149e6e7aeeb6ab0ad2,

title = "Domain-adaptive state of health prediction of vehicle batteries powered by deep learning",

abstract = "State of health (SOH) estimation of battery packs in electric vehicles (EVs) is essential for transportation electrification safety and reliability. The noise and complexity of EV battery pack data hinder the effectiveness of various data-driven SOH estimation methods using lab data. To address these challenges and achieve more effective data-driven EV battery pack SOH predictions, this study develops a comprehensive deep-learning-based SOH modeling framework for EV batteries. The framework begins with a two-stage mode decomposition (TSMD) method designed to effectively identify neat SOH degradation patterns better representing noisy field data. Next, an endogenous and exogenous multibranch network structure with a hierarchically fused attention mechanism (EEMB-HiFA) is developed for real-time prediction of EV battery pack SOH. Computational experiments leveraging datasets from seven EVs are conducted to validate the accuracy and adaptiveness of the proposed EEMB-HiFA. The results show that the EEMB-HiFA can achieve a 96.49% improvement in accuracy compared to strong benchmarks considered. {\textcopyright} 2025 The Author(s)",

keywords = "data feature fusion, data-driven model, deep learning, domain adaptation, electric vehicles, lithium-ion battery, signal denoising, state of health",

author = "Minghe Li and Zicheng Fei and Luoxiao Yang and Zijun Zhang and Kwok-Leung Tsui",

year = "2025",

month = apr,

day = "17",

doi = "10.1016/j.xcrp.2025.102550",

language = "English",

}

TY - JOUR

T1 - Domain-adaptive state of health prediction of vehicle batteries powered by deep learning

AU - Li, Minghe

AU - Fei, Zicheng

AU - Yang, Luoxiao

AU - Zhang, Zijun

AU - Tsui, Kwok-Leung

PY - 2025/4/17

Y1 - 2025/4/17

N2 - State of health (SOH) estimation of battery packs in electric vehicles (EVs) is essential for transportation electrification safety and reliability. The noise and complexity of EV battery pack data hinder the effectiveness of various data-driven SOH estimation methods using lab data. To address these challenges and achieve more effective data-driven EV battery pack SOH predictions, this study develops a comprehensive deep-learning-based SOH modeling framework for EV batteries. The framework begins with a two-stage mode decomposition (TSMD) method designed to effectively identify neat SOH degradation patterns better representing noisy field data. Next, an endogenous and exogenous multibranch network structure with a hierarchically fused attention mechanism (EEMB-HiFA) is developed for real-time prediction of EV battery pack SOH. Computational experiments leveraging datasets from seven EVs are conducted to validate the accuracy and adaptiveness of the proposed EEMB-HiFA. The results show that the EEMB-HiFA can achieve a 96.49% improvement in accuracy compared to strong benchmarks considered. © 2025 The Author(s)

AB - State of health (SOH) estimation of battery packs in electric vehicles (EVs) is essential for transportation electrification safety and reliability. The noise and complexity of EV battery pack data hinder the effectiveness of various data-driven SOH estimation methods using lab data. To address these challenges and achieve more effective data-driven EV battery pack SOH predictions, this study develops a comprehensive deep-learning-based SOH modeling framework for EV batteries. The framework begins with a two-stage mode decomposition (TSMD) method designed to effectively identify neat SOH degradation patterns better representing noisy field data. Next, an endogenous and exogenous multibranch network structure with a hierarchically fused attention mechanism (EEMB-HiFA) is developed for real-time prediction of EV battery pack SOH. Computational experiments leveraging datasets from seven EVs are conducted to validate the accuracy and adaptiveness of the proposed EEMB-HiFA. The results show that the EEMB-HiFA can achieve a 96.49% improvement in accuracy compared to strong benchmarks considered. © 2025 The Author(s)

KW - data feature fusion

KW - data-driven model

KW - deep learning

KW - domain adaptation

KW - electric vehicles

KW - lithium-ion battery

KW - signal denoising

KW - state of health

UR - http://www.scopus.com/inward/record.url?scp=105002909341&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-105002909341&origin=recordpage

U2 - 10.1016/j.xcrp.2025.102550

DO - 10.1016/j.xcrp.2025.102550

M3 - RGC 21 - Publication in refereed journal

SN - 2666-3864

JO - Cell Reports Physical Science

JF - Cell Reports Physical Science

M1 - 102550

ER -

Domain-adaptive state of health prediction of vehicle batteries powered by deep learning

Abstract

Funding

Research Keywords

UN SDGs

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

GRF: Data-driven Methods with Stronger Domain Generalizability for Machine Fault Diagnosis

ITF: Development of An AI-Powered Software System for Forecasting the Full Sequence of Day-ahead Wind Power

Cite this

Domain-adaptive state of health prediction of vehicle batteries powered by deep learning

Abstract

Funding

Research Keywords

UN SDGs

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Projects

GRF: Data-driven Methods with Stronger Domain Generalizability for Machine Fault Diagnosis

ITF: Development of An AI-Powered Software System for Forecasting the Full Sequence of Day-ahead Wind Power

Cite this