基于GA-IPSO-KPCA和变权组合模型的电动汽车充电方法  

Charging Method of Electric Vehicle Based on the GA-IPSO-KPCA and the Combined Model with Variable Weights

作　　者：傅莹颖 葛泉波 李春喜 崔向科 FU Yingying;GE Quanbo;LI Chunxi;CUI Xiangke(Logistics Engineering College,Shanghai Maritime University,Shanghai 201306,China;School of Automation,Nanjing University of Information Science and Technology,Nanjing 210044,China;Hangzhou Jiema Energy Technology Co.,Ltd.,Hangzhou 310000,China;School of Economics and Management,Beijing Jiaotong University,Beijing 102603,China)

机构地区：[1]上海海事大学物流工程学院,上海201306 [2]南京信息工程大学自动化学院,江苏南京210044 [3]杭州颉码能源科技有限公司,浙江杭州310000 [4]北京交通大学经济与管理学院,北京102603

出　　处：《控制工程》2024年第4期712-721,共10页Control Engineering of China

基　　金：国家自然科学基金资助项目(61803136)。

摘　　要：需求电压和需求电流是充电桩对电动汽车安全充电的重要依据。然而,随着电池的老化,电池管理系统的数据可能出现错误,使得电动汽车在充电时存在安全隐患。针对该问题,建立最小二乘支持向量机和深度置信网络的组合预测模型,提出一种基于变权组合模型的电动汽车充电方法。首先,针对数据掉线缺失问题,使用K均值和反距离加权方法对数据进行插值;然后,使用改进的混合核主成分分析算法对完整数据进行主成分提取,并使用改进粒子群优化算法自动确定混合核函数的权重。基于真实电动汽车数据的实验结果表明,所提方法能够准确地预测需求电压和需求电流,具有实际意义和可行性。Demand voltage and demand current are important bases for the safe charging of electric vehicles by charging piles.However,as the battery ages,there may be errors in the data of the battery management system,posing a safety hazard for electric vehicles during charging.To solve this problem,a combined prediction model of least square support vector machine and deep belief network is established,and a charging method based on the combined model with variable weights for electric vehicles is proposed.Firstly,to address the issue of data disconnection and missing,K-means and inverse distance weighting methods are used to interpolate the data.Then,an improved hybrid kernel principal component analysis algorithm is used to extract principal components from the complete data,and an improved particle swarm optimization algorithm is used to automatically determine the weights of the mixed kernel function.Experimental results based on real electric vehicle data show that the proposed method can accurately predict the demand voltage and demand current,which is of practical significance and feasibility.

关 键 词：充电安全 组合预测 粒子群优化算法 核主成分分析 深度置信网络 最小相对熵 

分 类 号：TP191[自动化与计算机技术—控制理论与控制工程]