摘要
锂离子电池的性能在循环充放电、环境温度变化、自身材料老化等情况下会不断退化,严重影响电池正常可靠运行。为了提高锂离子电池寿命预测的精准度,提出了一种基于麻雀搜索算法优化极限学习机的锂离子电池寿命预测模型。选用麻雀搜索算法SSA优化ELM的权值和阈值,既减小了由于ELM随机产生权值和阈值导致预测结果不准确和回归模型不稳定等缺点,又改善了全局搜索能力。实验采用NASA锂离子电池数据集,用均方根误差RMSE和相关指数R2作为评价标准,对所提模型(SSA-ELM预测模型)、BP神经网络模型和ELM模型的预测结果进行对比分析。实验结果表明,相比于传统神经网络模型,SSA-ELM算法泛化性能更好,预测精度更加准确可靠。
The lithium ion battery in the cycle of charge and discharge,environmental temperature change,its own material aging and other conditions will lead to its performance degradation,seriously affecting the normal and reliable operation of the battery.Therefore,it is important to study the residual service life(RUL)of lithium-ion batteries.In order to improve the accuracy of lithium-ion battery life prediction,a model based on sparrow search algorithm is proposed.This method uses the sparrow search algorithm SSA to optimize the weights and thresholds of ELM,which not only reduces the inaccurate prediction results and unstable regression model due to random ELM generating weights and thresholds,but also improves the global search ability.NASA lithium-ion battery data set,RMSE and correlation index R2 taken as the evaluation criteria,analyze the proposed model(SSA-ELM prediction model),BP neural network model and ELM model.Experimental results show that the SSA-ELM algorithm has better generalization performance and more accurate and reliable prediction accuracy compared with the traditional neural network model.
作者
郝锐
王海瑞
HAO Rui;WANG Hairui(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650504,Yunnan,China)
出处
《农业装备与车辆工程》
2023年第5期87-91,共5页
Agricultural Equipment & Vehicle Engineering
关键词
麻雀搜索算法
极限学习机
锂离子电池
寿命预测
sparrow search algorithm
extreme learning machine
lithium-ion battery
life prediction
