摘要
锂离子电池荷电状态(SOC)估算作为电池管理系统(BMS)的重要功能,是合理利用EV车辆电池组能量的前提。现有估算方法中,基于模型驱动法过程繁琐、计算量大;基于数据驱动法对数据要求高,估计精度差。针对以上问题,提出了基于双向GRU(BidiGRU)神经网络结合H∞滤波器的复合估算方法来估算锂离子电池的SOC。用不同温度下北京应力循环工况(BJDST)来训练BidiGRU神经网络模型,再使用联邦测试程序驾驶时间表(US06)来对模型进行测试。测试结果的均方根误差(RMSE)最小可达到2.05%,平均绝对误差(MAE)最小可达到1.79%。用H∞滤波器优化后的RMSE和MAE均可降低到0.17%以下,最低可达到0.11%。结果表明该方法可在不同温度和不同工况下对锂离子电池SOC做出实时估计,并能够达到较高的精度。
As an important function of the battery management system(BMS),the SOC estimation of lithium-ion batteries is a prerequisite for the rational use of the energy of the EV battery pack.Among the existing estimation methods,the process based on the model-driven method is cumbersome and the amount of calculation is large;the data-driven method has high data requirement and poor estimation accuracy.In response to the above problems,a composite estimation method based on bidirectional GRU(BidiGRU)neural network combined with H∞filter was proposed to estimate the SOC of lithium-ion battery.The BidiGRU neural network model was trained using the Beijing stress cycle conditions(BJDST)at different temperatures,and then the federal test program driving timetable(US06)was used to test the model.The minimum root mean square error(RMSE)of the test results can reach 2.05%,and the minimum average absolute error(MAE)can reach 1.79%.The RMSE and MAE after optimization with the H∞filter can be reduced to below 0.17%,and the lowest can reach 0.11%.The results show that the method can make real-time estimation of the SOC of lithium-ion battery under different temperatures and different working conditions,and can achieve higher accuracy.
作者
桂阳
周飞
杨文
陈星
李康群
GUI Yang;ZHOU Fei;YANG Wen;CHEN Xing;LI Kangqun(State Key Laboratory of Helicopter Transmission Technology,Nanjing University of Aeronautics and Astronautics,Jiangsu Nanjing 210016,China)
出处
《电源技术》
CAS
北大核心
2022年第4期384-389,共6页
Chinese Journal of Power Sources
