Spatio-Temporal Frequency Domain Analysis of PMU Data for Unsupervised Event Detection

We consider the unsupervised detection of power system events based on PMU measurements. We first present a convolutive dictionary model as a generative model for the short time Fourier transform (STFTs) of PMU data streams from multiple PMUs in the presence of a power system event. In the presence of an event, the STFTs from different PMUs display a common two-dimensional event signature at the event time—with its magnitude varying across PMUs, thus we develop a convolutive dictionary model to capture this unique spatiotemporal correlation. We leverage this convolutive dictionary model to formulate a binary hypothesis testing for event detection and implement a generalized likelihood ratio test to perform unsupervised event detection. The efficacy of the proposed detector is demonstrated using the experiments based on real life PMU data collected from a US Interconnection.