A method for detection of Low Frequency Oscillatory modes in power system for wide area monitoring system

The integration of renewable energy sources with power grids has amplified interactions among energy resources, causing Low-Frequency Oscillations (LFOs). The assessment of LFOs is very crucial in order to incorporate corrective measures to ensure the small signal stability of power system. The Total Least Square Estimation of Signal Parameters via Rotational Invariance Techniques (TLS-ESPRIT) method, known for precision in mode estimation, relies on prior knowledge of frequency components and loses accuracy at low signal-to-noise ratios (SNR). To overcome these limitations, this paper proposes an improved TLS-ESPRIT method utilizing a Singular Value (SV) ratio and a sequential partitioning technique for LFO modal parameter estimation. Initially, a SV ratio based low rank Hankel matrix filter is implemented to enhance noise resistance. Subsequently, a novel Model Order (MO) estimation technique is introduced using a convex combination of eigenvalue weightage to scale eigenvalue dominance in the trace of autocorrelation matrix and advanced medoid-based partitioning is explored to segregate the modes into signal and noise subspaces. In order to examine the efficacy of the suggested technique, comparison test of simulation results is conducted with some recently developed methods for synthetic signal, PMU data from a two area four machine system, practical probing data from the western electricity coordinating council, and oscillatory power data from the western system coordinated council 9 bus system and IEEE 39 bus system, simulated on real-time digital simulator.