电力系统自动化2025，Vol.49Issue(21)：108-119,12.DOI:10.7500/AEPS20250323006
计及因果关系增强的锂离子电池健康状态估计方法
State of Health Estimation Method for Lithium-ion Batteries Considering Causality Enhancement
摘要
Abstract
With the rapid development of new power system,energy storage technologies such as lithium-ion battery play a crucial role in integrating renewable energy sources with high penetration rates.Rapid and accurate state of health(SOH)estimation of lithium-ion battery helps enhance the safety and reliability of energy storage systems.However,traditional data-driven methods based on correlation analysis suffer from insufficient interpretability and struggle to ensure the accuracy and generalization of battery SOH estimation beyond the training samples.To address this issue,a SOH estimation method for lithium-ion batteries considering causality enhancement and based on bidirectional gated recurrent unit(BiGRU)network is proposed.First,a causality enhancement model for lithium-ion batteries based on causal-observation fusion inference is established.This model constructs multi-temporal nonlinear causality-enhanced features from batteries measurement data to improve the interpretability and in-distribution generalization of SOH estimation.Then,considering the multi-temporal nonlinear causality-enhanced features of the battery,the SOH estimation method for lithium-ion battery based on BiGRU network and invariant risk minimization theory is proposed to enhance the accuracy and out-distribution generalization of SOH estimation.Finally,the proposed method is validated using various actual operational condition data of lithium-ion batteries.Case study results demonstrate that the proposed SOH estimation method for lithium-ion batteries has good interpretability,accuracy,and generalization.关键词
新型电力系统/锂离子电池/健康状态/因果关系/不变风险最小化/储能Key words
new power system/lithium-ion battery/state of health(SOH)/causality/invariant risk minimization/energy storage引用本文复制引用
彭里卓,林达,汪湘晋,沈绍斐,高志海,杨莉,林振智..计及因果关系增强的锂离子电池健康状态估计方法[J].电力系统自动化,2025,49(21):108-119,12.基金项目
国家电网有限公司科技项目(5419-202319467A-3-2-ZN). This work is supported by State Grid Corporation of China(No.5419-202319467A-3-2-ZN). （5419-202319467A-3-2-ZN）
