摘要
针对锂离子电池健康状态(SOH)估算精度低的问题,提出一种基于主成分分析(PCA)与改进LM算法-双高斯核RBF(ILM-DGRBF)神经网络的方法,实现了SOH的准确估算。首先,提取与锂离子电池容量衰退高度相关的健康因子(HI),采用PCA方法进行降维处理,减少HI之间冗余度。其次,创建双高斯核RBF神经网络,利用改进LM算法实现网络参数在线学习,建立ILM-DGRBF神经网络。再次,利用数据增强的电池测试数据训练ILM-DGRBF实现SOH估算。验证表明,经PCA降维得到的主成分1能够有效地反应锂离子电池的老化趋势,可用于SOH的估算;与其他模型相比,所建ILM-DGRBF模型具有更高的估算精度和更好的鲁棒性,估算结果的误差控制在1.5%以内。最后,基于该方法构建一种新的SOH智能估算系统,为电池安全管理提供参考依据。
Aiming at the problem of low estimation accuracy of Li-ion battery state of health(SOH),a method based on principal component analysis(PCA)and improved Levenberg-Marquardt algorithm-double Gaussian kernel RBF(ILM-DGRBF)neural network was proposed,which realized the accurate estimation of SOH.Firstly,the health indicator(HI)highly related to the capacity decline was extracted,and PCA method was used for dimensional reduction processing to reduce the redundancy between HI.Secondly,a double Gaussian kernel RBF neural network was created,and improved LM algorithm was used to realize the online learning of neural network parameters to establish ILM-DGRBF neural network.Thirdly,ILM-DGRBF was trained with the enhanced battery test data to realize SOH estimation.The verification shows that the principal component 1 obtained by PCA dimensionality reduction can effectively reflect the aging trend of Li-ion battery,and can be used for SOH estimation;Compared with other models,the established ILM-DGRBF model has higher estimation accuracy and better robustness,and the error of the estimation results is controlled within 1.5%.Finally,based on this method,a new SOH intelligent estimation system was constructed to provide a reference basis for battery safety management.
作者
李亚飞
王泰华
张润雨
张家乐
Li Yafei;Wang Taihua;Zhang Runyu;Zhang Jiale(School of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo 454003,China)
出处
《电子测量技术》
北大核心
2023年第17期30-36,共7页
Electronic Measurement Technology
基金
国家自然科学基金(51974326)项目资助。
关键词
锂离子电池
健康状态
主成分分析
RBF神经网络
高斯核函数
LM算法
Li-ion battery
state of health
principal component analysis
RBF neural network
Gaussian kernel function
LM algorithm
