基于BP神经网络的锂电池组SOC估算  

SOC Estimation of Lithium Battery Pack Based on BP Neural Network

作　　者：贾海峰 李聪[1] Jia Haifeng;Li Cong(School of Automotive Engineering,Shanghai University of Engineering and Technology,Shanghai 201600,China)

机构地区：[1]上海工程技术大学汽车工程学院

出　　处：《农业装备与车辆工程》2020年第1期105-107,112,共4页Agricultural Equipment & Vehicle Engineering

摘　　要：电池荷电状态(SOC)的精确估计是电动车辆的核心技术之一,对影响电池荷电状态的因素进行分析归纳后,采用经典反向传播神经网络(BP神经网络)算法的动力电池SOC估计方法。利用高级车辆仿真软件ADVISOR对电动汽车典型行驶工况进行模拟,得到动力电池组电压、电流、平均温度和荷电状态数据,样本数据经归一化处理后导入神经网络模型中训练和测试,结果表明,该算法能有效提高SOC估算精度,具有较好的收敛性和鲁棒性,SOC估计误差范围能减小到4%以内,满足实际应用的需求。Accurate estimation of state of charge is one of the core technologies of electric vehicle.After analyzing and summarizing the influencing factors of the charge state of the battery,the SOC estimation method of power battery based on back propagation neural network(BP neural network)was proposed.The ADVISOR,an advanced vehicle simulation software,was used to simulate typical driving conditions of electric vehicles,the voltage,current,average temperature and state of charge data of the power battery pack were obtained.After normalization,the sample data were imported into the neural network model for training and testing.The results show that this algorithm can effectively provide SOC estimation accuracy and has good convergence and robustness.SOC estimation error range can be reduced to within 4%,which meets the need of practical application.

关 键 词：动力电池组 SOC 估算算法 预测精度 

分 类 号：TM912[电气工程—电力电子与电力传动] U469.72[机械工程—车辆工程]