Comprehensive Clustering of Disturbance Events Recorded by Phasor Measurement Units

Identifying disturbance events recorded by phasor measurement units (PMUs) has drawn researchers' attention in recent times. Some approaches to identify typical disturbance events frequently occurring in power systems have been documented. However, in order to comprehensively identify all disturbance events recorded by PMUs, it is required to know how many types of events are detected and recorded by PMUs monitoring a certain power system. In other words, for classification purposes, one must know the number of classes to begin with. This paper uses actual disturbance files stored in the database of a North American utility from 2007 through 2010 to determine how many classes (types of disturbances) the disturbance files contain. After analyzing various clustering techniques, a suitable unsupervised learning technique has been chosen and successfully implemented for this purpose. The results show that this process should be underpinned to any comprehensive event detection tool for PMU data.