基于PMU量测数据的噪声环境下电力系统故障检测  

Fault detection of power system in noise environment based on PMU measurement data

作　　者：黄赛 周毅博[1] HUANG Sai;ZHOU Yibo(School of Electrical Engineering,Northeast Electric Power UniversityNortheast Electric Power University,Jilin 132012,China)

机构地区：[1]东北电力大学电气工程学院,吉林吉林132012

出　　处：《电气应用》2023年第8期38-44,共7页Electrotechnical Application

基　　金：吉林省科技发展计划项目(20220208150RC)。

摘　　要：随着广域量测系统(Wide-Area Measurement System,WAMS)部署的日趋完善,为电力系统提供了海量高准确度的量测数据,这些数据被广泛用于电力系统的故障检测。但量测数据中噪声的存在会影响故障检测的准确性,导致漏判的发生,使电网面临故障范围扩大的风险。因此,提出了一种适用于噪声环境下的电网故障检测方法,该方法源于随机矩阵理论(Random Matrix Theory,RMT),通过对量测数据矩阵进行增维处理以弱化噪声的影响,并对数据源矩阵进行预处理,从理论上推导出抗噪性能优异的故障检测指标,通过比较故障检测指标与阈值的大小关系实现电网的故障检测,为调度运行人员实时监视电网运行状态提供依据。在IEEE 39节点系统设置算例,验证了该方法的有效性。With the increasing perfect deployment of wide-area measurement system,it provides massive high-precision measurement data for the power system,which is widely used for the fault detection of the power system.However,the existence of noise in the measurement data will affect the accuracy of fault detection,lead to the occurrence of leakage judgment,and make the power grid face the risk of expanding the fault range.Therefore,the paper proposed a grid fault detection method for noise environment,the method derived from random matrix theory,through the measurement data matrix to weaken the noise,and the matrix of data source,theoretically derived excellent fault detection index,by comparing the fault detection index and the threshold size relationship,provides the basis for dispatching operators real-time monitoring of power grid operation status.The effectiveness of the proposed method is verified in the IEEE39 node system setting examples.

关 键 词：故障检测 量测数据 噪声 随机矩阵理论 最大特征值 

分 类 号：TM73[电气工程—电力系统及自动化]