全文下载:
202506015.pdf
文章编号: 1672-6987(2025)06-0118-08 DOI: 10.16351/j.1672-6987.2025.06.015
方凤媛, 籍艳*(青岛科技大学 自动化与电子工程学院,山东 青岛 266061)
摘要: 精准辨识锂电池参数对荷电状态(state of charge,SOC)估计意义重大,SOC作为电池管理系统(battery management system,BMS)中最核心的技术之一,对电池系统的安全高效利用至关重要。本文基于锂电池的二阶RC等效电路模型,使用多新息随机梯度(multi innovation stochastic gradient,MISG)算法对锂电池参数进行了辨识,即将当前时刻的单个新息扩展为包含当前和过去时刻数据的多个新息来辨识参数,其目的是增加输入输出信息量,提高锂电池参数的精确度。同时,在不同实验仿真下,根据MISG算法和随机梯度算法(stochastic gradient,SG)得出的预测端电压值和实际端电压值的进一步对比,说明MISG算法可以增强参数的可靠性。并且,MISG算法估计的端电压平均绝对误差(mean absolute error,MAE)最大在2%左右,均方根误差(root mean square error,RMSE)最大在4%左右,相比于SG算法有更好地适应性和准确性,可以更好地改善参数估计精度。
关键词: 锂电池; 二阶RC等效电路模型; 多新息随机梯度; 参数辨识
中图分类号: TM 92 文献标志码: A
引用格式: 方凤媛, 籍艳. 新能源锂电池在线参数估计方法[J]. 青岛科技大学学报(自然科学版), 2025, 46(6): 118-125.
FANG Fengyuan, JI Yan. Online parameter estimation method of new energy lithium battery[J]. Journal of Qingdao University of Science and Technology(Natural Science Edition), 2025, 46(6): -.
Online Parameter Estimation Method of New Energy Lithium Battery
FANG Fengyuan, JI Yan(College of Automation and Electronic Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: Accurate identification parameters of lithium battery is great significance for the state of charge (SOC) estimation, the SOC is one of the core technologies in the battery management system (BMS), it is very important for the safe and efficient utilization of battery system. In this paper, presenting the second-order RC equivalent circuit model based on lithium battery, the multi-innovation stochastic gradient (MISG) algorithm is used to identify parameters of lithium battery, by extending a single innovation at current instant to a multi-innovation with current and past instant data to identify parameters, the aim is to increase the amount of input and output information, and improve the accuracy of lithium battery parameters. Meanwhile, according to the stochastic gradient (SG) algorithm and the MISG algorithm further comparison between the predicted terminal voltage and the actual terminal voltage under different experimental conditions, shows that the algorithm can enhance the reliability of the parameters. In addition, the maximum mean absolute error (MAE) estimated by MISG algorithm is about 2%, the maximum root mean square error (RMSE) is about 4%, compared with the SG algorithm, it has better adaptability and accuracy, and can improve the accuracy of parameter estimation better.
Key words: lithium-ion battery; second-order RC equivalent circuit model; multi-innovation stochastic gradient; parameter identification
收稿日期: 2024-12-24
基金项目: 国家自然科学基金项目( 62273167);山东省自然科学基金项目(ZR2025MS1104).
作者简介: 方凤媛 (1998—) , 女, 硕士研究生. * 通信联系人.