An Enhanced Prony Algorithm for On-line Detection of Small Signal Oscillations for Synchrophasor Application

This paper presents an online detection method of small signal oscillations using an enhanced Prony algorithm. The proposed method has considered the affect of missing measure-ments of phasor measurement units (PMUs) which occurs as a result of network congestion or defect in PMUs or phasor data concentrators (PDCs). In this context, at first, a sequential K nearest neighbours (SKNN) classifier is utilized to provide a robust data set to address such issue. In the second step, improved Prony algorithm is used to identify the oscillatory modes. The proposed approach has been compared to Matrix Pencil, Eigen Realization algorithm (ERA) and improved Prony for generated test signals with missing data at different noise levels. The suitability of the proposed monitoring scheme is further demonstrated on two area network and real PMU measurements derived from the Western Electricity Coordinating Council (WECC).