Real-Time Classification and Tracking of Wideband Synchronized Measurement Data Using Density Reachable Criterion

As renewable energy sources are increasingly integrated into the power grid, comprehensive perception and measurement of power system signals are essential for maintaining system stability. Wideband Measurement Units (WMUs) can synchronize measurements of frequency components (Fcs) accurately within the 0-10kHz range. However, the data always contain significant noise, with each measurement process conducted independently, leading to unknown modal affiliations of the data. To address this issue, this paper develops an online classification and identification method for wideband measurement data, which combines Mahalanobis and Euclidean distances as the density reachability criterion. The algorithm successfully validated with an actual dataset of wideband voltage and current measurements. This method effectively addresses the complexities introduced by the noise and independent measurement processes, ensuring more reliable applications at main control stations.