基于模糊RBF神经网络改进VSM的EV充放电控制策略研究  

Research on EV charging and discharging control strategy based on VSM improved by fuzzy RBF neural network

作　　者：马玉立 陈良亮 赵阳[2] MA Yuli;CHEN Liangliang;ZHAO Yang(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 210000,China;School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210000,China;NARI Technology Co.,Ltd.,Nanjing 210000,China)

机构地区：[1]南瑞集团(国网电力科学研究院)有限公司,江苏南京210000 [2]南京师范大学电气与自动化工程学院,江苏南京210000 [3]国电南瑞科技股份有限公司,江苏南京210000

出　　处：《现代电子技术》2023年第21期94-98,共5页Modern Electronics Technique

基　　金：国家重点研发计划项目(2021YFB2501605)。

摘　　要：为解决电动汽车(EV)大规模接入电网导致的电网频率、电压等指标稳定性下降的问题,提出将虚拟同步机(VSM)控制策略应用于EV充放电系统的前级双向AC/DC变换器中。针对VSM控制中虚拟惯量和虚拟阻尼参数整定难的问题,提出一种基于模糊RBF神经网络改进VSM的控制策略。首先,建立EV前级变换器VSM闭环传递函数,根据二阶系统稳定性分析虚拟参数对系统性能的影响;然后,融合模糊算法启发式搜索的优点和RBF神经网络优良的非线性函数逼近能力,设计模糊RBF神经网络自适应控制器,对VSM虚拟参数进行在线调整;最后,通过Matlab/Simulink仿真验证所提策略的有效性。In view of the power grid frequency and voltage stability degradation caused by large⁃scale access of electric vehicle(EV)to the power grid,this paper proposes that the virtual synchronous machine(VSM)control strategy can be applied to the front⁃stage bidirectional AC/DC converter of the EV charging and discharging system.It is difficult to tune the virtual inertia and virtual damping parameters in VSM control,so a control strategy based on VSM improved by fuzzy RBF(radial basis function)neural network is proposed.The VSM closed⁃loop transfer function of the EV front⁃end converter is established,and the influence of virtual parameters on system performance is analyzed based on the second⁃order system stability.Then,combining the advantages of heuristic search of fuzzy algorithm with the excellent nonlinear function approximation ability of RBF neural network,the fuzzy RBF neural network adaptive controller is designed to adjust the virtual parameters of VSM online.The effectiveness of the proposed strategy is verified on Matlab/Simulink.

关 键 词：EV充放电 虚拟同步机 模糊控制 RBF神经网络 自适应控制 模糊算法 

分 类 号：TN876-34[电子电信—信息与通信工程]