학술논문
Comparative Assessment of Prony Analysis and Eigensystem Realization Algorithm for Forced Oscillation Detection and Mode Estimation Considering PMU Noise
Document Type
Conference
Source
2023 IEEE 3rd International Conference on Smart Technologies for Power, Energy and Control (STPEC) Smart Technologies for Power, Energy and Control (STPEC), 2023 IEEE 3rd International Conference on. :1-6 Dec, 2023
Subject
Language
Abstract
In this study, Prony and Eigensystem Realization Algorithm (ERA) are compared for identifying electromechanical and forced oscillation modes. The varying number of excited modes in practical power systems makes the process of determining model order challenging. Prony requires an iterative procedure, while ERA utilizes Singular Value Decomposition (SVD) for direct model order determination. Comparitive studies shows that ERA accurately estimates the signal, and generates fewer insignificant modes compared to Prony. Identified modes are used to design a filter, which estimates desired signal frequency. The filtered signal’s periodogram is then compared with detection threshold to detect Forced oscillation (FO) at specific frequencies.