基于SOA-BP联合算法的锂离子电池SOC估计研究  

Research on SOC Estimation of Lithium-Ion Batteries Based on SOA-BP Joint Algorithm

作　　者：于仲安 王涛 肖泽锴 YU Zhongan;WANG Tao;XIAO Zekai(School of Electrical Engineering and Automation,Jiangxi University of Science and Technology,Ganzhou,Jiangxi 341000,China;Institute of international Education,Nanchang Institute of Technology,Nanchang,Jiangxi 330000,China)

机构地区：[1]江西理工大学电气工程与自动化学院,江西赣州341000 [2]南昌工程学院国际教育学院,江西南昌330000

出　　处：《自动化应用》2025年第3期173-176,179,共5页Automation Application

基　　金：国家自然科学基金(52167005);江西省研究生创新专项资金项目(YC2023-S617)。

摘　　要：为确保电动汽车安全稳定运行,需准确估计电池荷电状态(SOC)。针对单一BP神经网络在SOC估计中精确度不高的问题,提出了一种基于改进海鸥优化算法(SOA)和BP神经网络的锂离子电池SOC估计方法。通过改进SOA优化BP神经网络的初始权值和阈值,克服单一BP神经网络易陷入局部极小值和收敛性不佳等缺点。经仿真实验表明,与GA-BP和单一BP神经网络预测相比,SOA-BP神经网络在锂离子电池SOC估计中平均绝对误差分别减少了1.1%和4.53%,具有更高的精准度和收敛性。To ensure the safe and stable operation of electric vehicles,it is necessary to accurately estimate the State of Charge(SOC)of the battery.A lithium-ion battery SOC estimation method based on improved Seagull Optimization Algorithm(SOA)and BP neural network is proposed to address the issue of low accuracy in SOC estimation using a single BP neural network.By improving SOA to optimize the initial weights and thresholds of BP neural networks,the shortcomings of a single BP neural network,such as being prone to local minima and poor convergence,can be overcome.Simulation experiments have shown that compared with GA-BP and single BP neural network prediction,SOA-BP neural network reduces the average absolute error in lithium-ion battery SOC estimation by 1.1%and 4.53%,respectively,and has higher accuracy and convergence.

关 键 词：海鸥优化算法 BP神经网络 锂离子电池 荷电状态 

分 类 号：TM912[电气工程—电力电子与电力传动]