摘要
针对电动汽车电池荷电状态(SOC)估算不准确的问题,采用列文伯格—马夸尔特(LM)算法搭建了电池SOC动态预测模型,并充分考虑了电池电压、电流、温度等影响因素。采集系统通过串口与LabVIEW监测系统进行通信,实现了电池数据的分析。利用MATLAB将采集的电池数据用于模型实验,实验结果表明:利用搭建的模型预测电池的SOC提高了预测精度,具有普适性。
Aiming at problem that estimation of battery state of charge( SOC) electric vehicles is inaccurate,Levenberg-Marquardt( LM) algorithm is used to build dynamic prediction model for battery SOC,and the influencing factors such as battery voltage,current and temperature are considered. Acquisition system communicate with LabVIEW monitoring system through serial port to achieve battery data analysis. Collected battery data is used for model experiment by MATLAB. Experimental results show that using the built model to predict battery SOC improves prediction precision,it has universality.
作者
李桂娟
张持健
施志刚
李亮
刘雪
LI Gui-juan;ZHANG Chi-jian;SHI Zhi-gang;LI Liang;LIU Xue(School of Physics and Electronic Information,Anhni Normal University,Wuhu 241002,China)
出处
《传感器与微系统》
CSCD
2018年第10期69-71,共3页
Transducer and Microsystem Technologies
关键词
锂电池
LABVIEW
荷电状态预测
反向传播神经网络
lithium battery
LabVIEW
state of charge(SOC)prediction
back propogation neural network (BPNN)