基于MEA-BP神经网络的电池储能系统SOC状态评估  

State of Charge Estimation for Battery Energy Storage System Based on MEA-BP Neural Network

作　　者：孙威 修晓青[2] 肖海伟 余晓玲 张亮 刘明爽 

机构地区：[1]深圳市欣旺达综合能源服务有限公司,广东深圳518100 [2]中国电力科学研究院,北京100192

出　　处：《电器与能效管理技术》2018年第1期51-54,83,共5页Electrical & Energy Management Technology

基　　金：山西省重点研发计划重点项目(201603D112001);国家电网公司总部科技项目(DG71-17-012)

摘　　要：提出了一种基于MEA-BP神经网络的电池储能系统荷电状态(SOC)预测方法,可准确评估储能电池的SOC状态。选择较易获得的电池外特性电流、电压和温度作为BP神经网络的输入,SOC状态作为神经网络输出,并利用思维进化算法(MEA)优化网络结构,对电池储能系统SOC状态进行评估。将评估结果与传统BP神经网络评估结果进行对比,结果表明,该方法可实现对电池储能系统SOC的准确评估,且预测误差小,具有良好的工程应用前景。A novel method of the SOC estimation for the battery energy storage system（BESS） based on MEA- BP neural network was proposed. The external characteristics of the battery is selected as the input of BP neural network,the operating status of the SOC is regarded as the output, the weights of the BP neural network are optimized by the mind evolutionary algorithm（MEA）. The simulation results show that compared with the traditional BP neural network,MEA-BP network has higher accuracy and has a broad prospect of engineering application.

关 键 词：电池储能系统 BP神经网络 荷电状态(SOC) MEA算法 

分 类 号：TM912[电气工程—电力电子与电力传动]