基于随机矩阵理论的配电网运行状态相关性分析方法  

A Correlation Analysis Method for Operation Status of Distribution Network Based on Random Matrix Theory

作　　者：徐心怡[1] 贺兴[1] 艾芊[1] 邱才明 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海市闵行区200240 [2]上海交通大学大数据工程技术研究中心,上海市闵行区200240

出　　处：《电网技术》2016年第3期781-790,共10页Power System Technology

基　　金：国家自然科学基金项目(51577115;61571296)~~

摘　　要：配电网的运行状态是各种影响因素共同作用的结果。为了挖掘各种影响因素与配电网运行状态之间的内在联系,提出一种基于随机矩阵理论的相关性分析方法。首先提出一种增广矩阵法,利用影响因素数据和电网运行状态数据构造数据源矩阵。然后基于实时分离窗技术和随机矩阵理论,分析数据源矩阵中元素的统计特性,并将分析结果与随机矩阵理论的理论预测进行比较。在分析过程中,采用线性特征值统计量(如平均谱半径)作为相关性指标。算例分析表明,该方法能够实时、定量地分析复杂系统中海量数据的相关性,揭示1种或多种影响因素对电网运行状态之间的影响。The operating status of distribution network is the interaction result of various factors. In order to explore inherent relationships between various factors and operating status of distribution network, a correlation analysis method is proposed based on random matrix theory. As a data source, an augmented matrix is formulated by two parts： status data and factor data. Based on real-time split-window and random matrix theory, analysis results of data correlations are obtained and compared with theoretical predictions. In this procedure, the linear eigenvalue statistic, such asmean spectral radius, is calculated as correlation indicator. Case studies validate performance of the proposed method on analyzing correlations of massive data in complex systems in real time, revealing influences of one or more factors on operating status of distribution network.

关 键 词：配电网 相关性分析 随机矩阵理论 增广矩阵 线性特征值统计量 平均谱半径 

分 类 号：TM71[电气工程—电力系统及自动化]