基于数据驱动的锂离子电池荷电状态估计
潘凌¹，汪振东²
（1. 国网乌鲁木齐供电公司营销服务中心，新疆 乌鲁木齐 830022；2. 国网新疆电力公司营销部，新疆 乌鲁木齐 830022）
摘要： 针对锂离子电池荷电状态不可以直接测量的问题，提出一种通过测量电化学阻抗并采用机器学习算法进行估
计的方案。根据锂离子电池阻抗特性，控制温度一定，选取特定频率下阻抗的实部、虚部和相角正切值作为特征参量，
通过线性回归和BP神经网络的方法对数据进行训练并估计荷电状态，进行误差分析。
关键词： 锂离子电池；机器学习；电化学阻抗；荷电状态预测
中图分类号： TM912     文献标志码： A     文章编号： 2095-0802-(2022)07-0041-03
State of Charge Estimation of Lithium Ion Batteries Based on Data Driven
PAN Ling¹, WANG Zhendong²
(1. Marketing Service Center, State Grid Urumqi Power Supply Company, Urumqi 830022, Xinjiang, China; 2. Marketing
Department, State Grid Xinjiang Electric Power Company, Urumqi 830022, Xinjiang, China)
Abstract: Aiming at the problem that the state of charge of lithium ion batteries can not be measured directly, an estimation
scheme by measuring electrochemical impedance using machine learning algorithm was proposed. According to the impedance
characteristics of lithium ion batteries, the control temperature was fixed, and the real part, imaginary part and tangent value of
phase angle of impedance under specific frequency were selected as characteristic parameters, the data were trained and the state
of charge was estimated by linear regression and BP neural network, and the error was analyzed.

Key words: lithium ion battery; machine learning; electrochemical impedance; state of charge estimation