基于改进型LMBP神经网络方法对蓄电池荷电状态的预测  

Prediction of the state of charge of battery based on improved LMBP neural network method

作　　者：王筱璇 侯冠军 孙思豪 WANG Xiaoxuan;HOU Guanjun;SUN Sihao(Xilingol Ultra High Voltage Power Supply Bureau,Inner Mongolia Electric Power(group)Co.,Ltd.,Xilinhot Inner Mongolia 026000,China)

机构地区：[1]内蒙古电力(集团)有限责任公司锡林郭勒超高压供电局,内蒙古锡林浩特026000

出　　处：《蓄电池》2018年第2期69-72,共4页Chinese LABAT Man

摘　　要：针对变电站中因过度充放电而导致阀控式铅酸蓄电池寿命短且利用率低的问题,通过改进型LMBP神经网络模型对铅酸蓄电池的荷电状态(SOC)进行预测,能够加快计算速度与精度,有效提升蓄电池的寿命与使用率。在Matlab环境下对铅酸蓄电池放电过程进行仿真研究的结果验证了,改进型LMBP型神经网络算法能有效提高SOC的估算精度,延长电池寿命。An improved LMBP neural network model is used to predict the state of charge(SOC)of lead-acid batteries for solving the problems of short lifespan and low utilization rate of the valveregulated lead-acid batteries caused by excessive charging and discharging in the substations.This model could improve the computing speed and precision,extend the life of batteries and increase the utilization rate of batteries.In addition,a simulation of the discharge process is conducted using Matlab,and the results prove that the improved LMBP neural network model can effectively improve the estimation precision of the SOC and extend the lifespan of batteries.

关 键 词：变电站 蓄电池 LMBP 神经网络 荷电状态(SOC) 估算精度 

分 类 号：TM912.9[电气工程—电力电子与电力传动]