A Novel Method for Screening the PMU Phase Angle Difference Data Based on Hyperplane Clustering

Compared with the traditional supervisory control and data acquisition (SCADA) data, phasor measurement unit (PMU) data is characterized by phase angle measurement and high reporting speed (perhaps 100 Hz). The high reporting speed provides dynamic characteristics of the power system frequency, voltage, and current measurement. PMUs have become one of the important data sources for smart grid monitoring. PMU/WAMS (wide area measurement system) based advanced applications have been widely used in the dispatch centers. Some of the applications, such as line parameter identification and state estimation, depend not only on phase angle data but also on phase angle difference between different locations. Field data can suffer from errors, such as time synchronization error, transducer error, PMU algorithm error, hardware error or malicious attacks, etc. A time synchronization error can directly cause an error in the phase angle difference calculated between the two ends of a transmission line that could degrade a PMU based application. In this paper, a novel method to cluster the phase angle difference data, assess the data quality and screen out the bad PMU phase angle difference data is proposed. First, we develop the hyperplane cluster method to cluster the phase angle difference data. Second, in order to screen out the right data type, this paper compares the virtual reactance parameters of each data type obtained by voltage mean to the line reactance parameter given by the system model. Finally, the performance of the proposed methods has been verified by a simulation. The efficiency of the proposed method has been analyzed. The application of the proposed method using field measured PMU data shows the engineering practicability of the proposed method. In addition, the comparison of the proposed method with other clustering methods is discussed.