Using sparsity to estimate oscillatory mode from ambient data

Abstract

In this paper, a sparse-technique-based representation of the signal over a learned dictionary and random decrement technique are explored to extract the oscillatory mode from the ambient data. The main contribution of the present work is to design a dictionary and compute the coefficients that best represent the clean signal to estimate the modes. In this work, the noise embedded in the ambient signal is minimized by representing the ambient signal in sparse domain with respect to the dictionary. Comparison between the proposed method and other methods such as nonlinear filtering, etc., has been done on the test signal, two-area power system on the data generated through simulation in Matlab, two-area data simulated on real-time digital simulator and real measurement from Phasor data concentrator (PDC) of Indian power system and Western Electricity Coordinating Council (WECC) network.