基于随机矩阵的电网多元数据异常状态诊断  

Abnormal State Diagnosis of Power Grid Multivariate Data Based on Random Matrix

作　　者：杨扬 刘道杰 李炜[2] YANG Yang;LIU Daojie;LI Wei(Shandong Institute of Petrochemical Engineering,School of Mechanical and Control Engineering,Dongying 257061,China;Technical Testing Center of Shengli Oilfield Company,Dongying 257061,China)

机构地区：[1]山东石油化工学院机械与控制工程学院,山东东营257061 [2]中国石油化工股份有限公司胜利油田分公司,山东东营257061

出　　处：《计算机测量与控制》2023年第10期40-48,共9页Computer Measurement &Control

基　　金：东营市科学发展基金项目(DJ2020014)。

摘　　要：为了解决电网故障选线难题,有效提升电网安全运行水平,提出了一种基于随机矩阵系列理论的电网双重判据定位诊断方法,用于电网异常检测与故障类型识别;首先通过借助PMU采集电网各节点处三相电压、三相电流数据构建样本数据源矩阵,再分别从时间维度和空间维度判断故障时间、故障馈线及故障类型,实现纵横多重分析;其中时间维度以谱偏离度和平均谱半径为指标,M-P律和单环定理作为双重判据确定故障时间;空间维度采用滑动时窗得出样本协方差矩阵进行故障位置判定和故障类型识别;最后借助Matlab软件,以IEEE39节点系统和某油田实际电网为例,进行电网故障时刻确定、故障位置判定和故障类型识别;实验结果证明该方法相比其他算法,优势在于谱偏离度的大小比之前小了很多,从而大大扩展了故障检测的范围,提高了故障识别的效率。In order to solve the problem of power grid fault line selection and effectively improve the safe operation level of power grid,a double criterion location diagnosis method based on random matrix series theory is proposed for power grid anomaly detection and fault type identification.Firstly,the sample data source matrix is constructed by collecting three-phase voltage and three-phase current data at each node of the power grid with the help of PMU,and then the fault time,fault feeder and fault type are judged from time latitude and space latitude respectively to realize vertical and horizontal multiple analysis.Among them,the time latitude is determined by the spectral deviation and the average spectral radius,and the M-P law and the single-loop theorem are used as double criteria to determine the fault time.In the spatial dimension,the sample covariance matrix obtained by sliding time window is used to determine the fault location and identify the fault type.Finally,with the help of MATLAB software,taking IEEE39-bus system and an actual power grid in an oil field as examples,the fault time,fault location and fault type identification of power grid are carried out.The experimental results show that compared with other algorithms,the advantage of this method is that the magnitude of spectral deviation is much smaller before,thus greatly expanding the scope of fault detection and improving the efficiency of fault identification.

关 键 词：高维随机矩阵 平均谱半径 谱偏离度 M-P定律 单环定理 

分 类 号：TM721[电气工程—电力系统及自动化]