Distribution system event detection and classification using local voltage measurements

Wide-Area Measurement Systems (WAMS) are being implemented in order to help increase situational awareness within the electric grid. These systems use Phasor Measurement Units (PMUs), devices which can measure the voltage and current within the power system. One kind of PMU is the Frequency Disturbance Recorder (FDR), which measures voltage magnitude, frequency, and phase angle at 10 data samples per second. These measurements are taken at the 120V level, resulting in a relatively low-cost, rapidly deployable alternative to other PMUs. This paper presents an approach to distribution system event detection and classification using voltage data obtained from FDRs installed around the University of Illinois at Urbana-Champaign (UIUC). This model-free classification of events will use pattern recognition techniques to help identify features that may be unique to these disturbances. Analysis will be applied to a sliding window of voltage data, and the results from each window are compared against one another in order to help determine what kind of event occurred.