基于随机矩阵理论的故障时刻确定和故障区域定位方法

A Random Matrix Theory-based Approach to Fault Time Determination and Fault Area Location

针对目前基于广域测量系统（wide-area measurement system,WAMS）数据的故障诊断方法容易受到WAMS不良数据干扰的问题,该文将随机矩阵理论应用于WAMS量测数据的信息提取上,分别提出了基于平均谱半径（mean spectral radius,MSR）指标的故障时刻确定方法和基于相量测量单元（phasor measurement units,PMU）数据相关性分析的故障区域定位方法,通过PMU数据的随机矩阵模型的建立和平均谱半径指标及PMU数据相关性的计算,实现了基于MSR指标的故障时刻确定和基于PMU数据相关性分析的故障区域定位,且对不良数据具有非常好的鲁棒性。分别以10机39节点算例和某实际区域电力系统为例,仿真验证了所提方法的有效性和准确性。

It is a problem that the fault diagnosis process using wide-area measurement system（WAMS） measurement data is susceptible to bad data.In this paper,the random matrix theory was applied to the information extraction of WAMS measurement data.The fault time determination method based on mean spectral radius（MSR） index and the fault location method based on phase measurement units（PMU） data correlation analysis were proposed,respectively.Through the modeling of the random matrix of PMU data and the mean spectral radius index and the PMU data correlation calculation,the fault time determination based on the MSR index and the fault area location based on the PMU data correlation analysis are realized,which can effectively avoid the interference of bad data.The validity and practicability of the proposed methods are verified by the simulation results of IEEE 39 benchmark system and actual WAMS data.