18650动力电池SOC的建模与估算研究  

Modeling and estimation study on 18650 power battery SOC

作　　者：王增红[1] 潘勇[1] 李旭军[1] 雷维新[1] 潘俊安 

机构地区：[1]湘潭大学材料与光电物理学院,湖南湘潭411105

出　　处：《电源技术》2014年第10期1809-1811,共3页Chinese Journal of Power Sources

基　　金：湖南省科技厅项目(2012GK4008)

摘　　要：为了实现对18650动力电池荷电状态(SOC)的准确在线估算,将外界条件划分为12种情况,研究了有效电量、电池衰老及环境温度对电池SOC的影响。建立了一个BP神经网络模型,在12种条件下对其进行样本学习训练,建立了一种能够适用于不同条件下的基于条件查找方法的SOC估算模型。通过与传统方法对比,证明改进的估算方法具有更好的精度和实用性。In order to achieve accurate on-line estimates on state of charge （SOC） of 18650 power battery, the effective power, cell aging and temperature influence were studied, and then the ambient conditions were divided into 12 kinds of situations. Meanwhile, a BP （ Back Propagation ） neural network model was established by training at 12 different kinds of conditions, and a conditional search based SOC estimation model was established to adapt to different measurement conditions. The improved estimation method was of better accuracy and practicability compared with the traditional method.

关 键 词：18650动力电池 荷电状态 BP神经网络 条件查找 有效电量 

分 类 号：TM912[电气工程—电力电子与电力传动]