基于模糊最小二乘支持向量机的微电网群状态评估方法  

Evaluation Method for Microgrid Cluster State Based on Fuzzy Least Squares Support Vector Machine

作　　者：陈卫东 梁朔 肖园园 郭敏 CHEN Weidong;LIANG Shuo;XIAO Yuanyuan;GUO Min(Electric Power Research Institute of Guangxi Power Grid Company Limited,Nanning 530023,China)

机构地区：[1]广西电网有限责任公司电力科学研究院,广西壮族自治区南宁市530023

出　　处：《电力系统自动化》2019年第2期76-82,共7页Automation of Electric Power Systems

基　　金：中国南方电网有限责任公司科技项目(GXKJXM20160317);国家高技术研究发展计划(863计划)资助项目(2012AA050217)~~

摘　　要：针对微电网群能量管理与协调控制系统适应多微电网间多工况控制策略灵活调整的需要,提出了一种基于模糊最小二乘支持向量机(FLS-SVM)的低压微电网群运行状态实时评估模型。该模型基于传统电力系统运行状态描述方法,建立了微电网群及子微电网安全正常运行的边界条件,以电压偏移率、储能剩余容量及充放电时间、发用电功率等多维度特征向量对子微电网状态分类,应用FLS-SVM对子微电网的实时运行状态进行评估,最后判别出微电网群运行状态。实例计算分析表明,该模型可跟随系统采样周期实时评估,对离、并网条件下子微电网运行状态均能实现准确有效地分类,为微电网群快速判断网内状态并灵活调整控制策略提供依据。Aiming at energy management system(EMS)and coordination control system(CCS)for adapting to flexible adjustment of microgrid cluster control strategy at variable operation states,a real-time operation state evaluation model of lowvoltage microgrid cluster based on fuzzy least squares support vector machine(FLS-SVM)is proposed.With traditional power system operation state description method,the proposed model builds constraints in secure and normal operation of microgrid cluster and microgrid.Microgrid states are classified in view of multi-dimensional eigenvector including voltage deviation rate,energy storage state of charge,energy storage charge and discharge time,power generation and load.The real-time operation states of microgrids are evaluated with FLS-SVM,and then the operation states of microgrid cluster can be determined.Case study shows that the method can conduct the real-time evaluation within the sampling period of the system,and classify the operation states of microgrids accurately and effectively under the conditions of both off-grid and on-grid.It can provide the basis for microgrid cluster to quickly determine the operation state and flexibly adjust control strategy.

关 键 词：微电网群 实时评估 最小二乘支持向量机 运行状态评估 

分 类 号：TM73[电气工程—电力系统及自动化]