摘要
锂离子电池的健康状态(SOH)对维持新能源电动汽车系统的稳定性至关重要。为提高锂电池SOH预测精度,提出一种基于变分模态分解(VMD)的猎人猎物优化(HPO)的神经基扩展分析(NBEATS)神经网络的SOH预测方法。首先,通过对电池老化数据的分析,提取与SOH高度相关的健康因子(HIs)并进行融合;其次,利用VMD方法将融合HI分解为多个模态分量,并使用HPO超参数优化的NBEATS模型来捕捉各模态分量的特征和时序规律。最终,通过加和重构各个分量的预测值来获得电池的SOH预测。在NASA电池数据集上的实验表明,与NBEATS、HPO-NBEATS和VMD-NBEATS模型相比,VMD-HPO-NBEATS模型在MAE、RMSE和r2评价指标上均有超2%的提升,证明所提方法在SOH预测的有效性与优越性。
The State of Health(SOH)of lithium-ion batteries is crucial for maintaining the stability of new energy electric vehicle systems.To improve the accuracy of SOH prediction for lithium batteries,a SOH prediction method based on the Variational Mode Decomposition(VMD)Hunter-Prey Optimization(HPO)Neural Basis Expansion Analysis(NBEATS)neural network is proposed.Firstly,by analyzing the battery aging data,health indicators(HIs)highly related to SOH are extracted and fused;secondly,the fused HIs are decomposed into multiple modal components using the VMD method,and the characteristics and temporal patterns of each modal component are captured using the NBEATS model with HPO hyperparameter optimization.Finally,the SOH prediction of the battery is obtained by summing and reconstructing the predicted values of each component.Ablation experiments on the NASA battery dataset show that compared with the NBEATS,HPO-NBEATS,and VMD-NBEATS models,the VMD-HPO-NBEATS model has an improvement of more than 2%in MAE,RMSE,and r2 evaluation metrics,proving the effectiveness and superiority of the proposed method in SOH prediction.
作者
李泽龙
乔钢柱
崔方舒
蔡江辉
史元浩
王博辉
LI Zelong;QIAO Gangzhu;CUI Fangshu;CAI Jianghui;SHI Yuanhao;WANG Bohui(School of Computer Science and Technology,North University of China,Taiyuan 030051,China;School of Electrical and Control Engineering,North University of China,Taiyuan 030051,China;School of Cyber Science and Engineering,Xi'an Jiaotong University,Xi'an 710049,China)
出处
《中国测试》
CAS
北大核心
2024年第9期65-73,共9页
China Measurement & Test
基金
山西省基础研究计划资助项目(202303021222084)
山西省基础研究计划联合资助项目(TZLH20230818007)
山西省研究生教育创新项目(2024KY613)。
作者简介
李泽龙(2001-),男,山西忻州市人,硕士研究生,专业方向为储能电池健康管理;通信作者:乔钢柱(1975-),男,陕西汉阴县人,教授,博士,研究方向为物联网技术及应用、大数据处理技术、区块链技术。
