电器与能效管理技术 ›› 2023, Vol. 0 ›› Issue (9): 55-61.doi: 10.16628/j.cnki.2095-8188.2023.09.009

• 估计与预测技术 • 上一篇    下一篇

基于GRA-MEA-BP算法的锂电池SOC估计方法研究

许守平1,2, 胡娟1,2, 付珊珊1, 陈娟1   

  1. 1.中国电力科学研究院有限公司,北京 100192
    2.新能源与储能运行控制国家重点实验室,北京 100192
  • 收稿日期:2022-12-22 出版日期:2023-09-30 发布日期:2023-11-23
  • 作者简介:许守平(1978—),男,高级工程师,主要从事锂电池储能、电力电子技术、新能源发电等研究。|胡 娟(1978—),女,高级工程师,主要从事锂电池储能、电力电子技术、灵活交流输电等研究。|付珊珊(1986—),女,高级工程师,博士,主要从事锂电池管理系统、电池检测、电力储能等研究。
  • 基金资助:
    国家电网公司科技项目(5442DG190013)

Research on SOC Estimation of Lithium Battery Based on GRA-MEA-BP Algorithm

XU Shouping1,2, HU Juan1,2, FU Shanshan1, CHEN Juan1   

  1. 1. China Electric Power Research Institute, Beijing 100192, China
    2. State Key Laboratory of Operation and Control of Renewable Energy & Storage Systems, Beijing 100192, China
  • Received:2022-12-22 Online:2023-09-30 Published:2023-11-23

摘要:

针对锂电池荷电状态(SOC)的非线性特性,提出GRA-MEA-BP神经网络算法,用于锂电池运行的实时SOC估计。所提方法通过灰度关联分析(GRA)筛选SOC估计的主要影响因子,将得到主要影响因子作为输入,当前时刻的SOC作为输出,训练MEA-BP神经网络。最后,以实际使用的锂电池作为对象,对所提出的方法进行数据训练和验证。结果表明了所提方法的准确性。

关键词: 锂电池, 荷电状态, 神经网络, 灰度关联分析

Abstract:

In view of the nonlinear characteristics of lithium battery state of charge (SOC),the grey relational analysis,mind evolutionary algorithm,back propagation (GRA-MEA-BP) neural network algorithm is proposed,which is used for the SOC estimation of lithium battery at the current moment.Several main impact factors of SOC estimation is selected by GRA.The MEA-BP neural network is trained by taking the main influencing factors as inputs and the current SOC as output.Finally,taking the actual lithium battery as the object the proposed method is trained and verified by real data.The verification results indicate the accuracy of the proposed method.

Key words: lithium battery, state of charge, neural network, gray correlation analysis

中图分类号: