A Systemic Deep Learning Approach to Battery SOH Estimation: Integrating IC Curve Analysis and STAR-Mamba Modeling with Real-World Driving Data
作者
Jianle Zheng,Peiwen Zhong,M. Lino da Silva,Hongchi Zhou,Yang Song,Yu Sheng Shi
出处
期刊: [Institute of Electrical and Electronics Engineers] 日期:2025-09-19卷期号:: 984-989
标识
DOI:10.1109/icpre67300.2025.11273926
摘要
The state of health (SOH) estimation of lithium-ion batteries in real-world electric vehicles is challenged by incomplete and inconsistent charging data. This paper proposes a method that leverages Gaussian Process Regression (GPR) to predict missing segments of Incremental Capacity (IC) curves and applies Dynamic Time Warping (DTW) to quantify temporal dissimilarities for monthly IC curve fusion. The fused features are then used to train a deep learning model for accurate SOH estimation. To further enhance performance, an improved model named STAR-Mamba is developed. Experimental results demonstrate that the proposed method maintains high accuracy under noisy and incomplete data conditions, achieving a MAPE of 1.01% and an RMSPE of 1.40%.