基于神经网络的电动汽车电池SOC估算研究

论文首先分析了当前电动汽车电池管理系统中存在的问题,特别是电池电压的精确测量和剩余电量的准确预测问题一直亟待突破,因此,论文在分析电池荷电状态(SOC)影响因素的基础上,进行了动力电池的充放电实验,建立了BP网络电池模型,通过对网络进行训练,应用神经网络模型进行SOC估算,实验表明:建立的BP网络具有较好的适应性,能有效预测锂离子动力电池电压、电流和放电容量间的映射关系。可以准确地对电动汽车电池进行SOC估算。

Research on SOC Estimation of Electric Vehicle Batteries Based on Neural Network

This paper first analyzes the existing electric vehicle battery management system problems,especially the problem of accurate measurement and accurate prediction of remaining battery voltage has been breakthrough,therefore,based on the analysis of the battery state of charge(SOC)based on the factors,the charge discharge test of power battery is carried out,BP network battery model is established,through the network training,SOC is estimated by neural network model,the experiments show that BP network has better adaptability and can effectively predict the mapping relationship between lithium ion power battery voltage,current and discharge capacity. It can accurately estimate SOC for EV batteries.