Thesis
Synchrophasor based ambient oscillation monitoring using short analysis window
Washington State University
Master of Science (MS), Washington State University
12/2015
Handle:
https://hdl.handle.net/2376/100124
Abstract
Power system oscillation monitoring plays a significant role for the safe operation of the system. Serious blackouts may be avoided if an effective method of small signal analysis is utilized. Traditional approach to analyze poorly damped low frequency electromechanical mode in power system is based on system model. However the linearized time-invariant model which is obtained around the equilibrium point is no longer valid when there are some small or large disturbance in the system. With the increasing installation of Phasor Measurement Units (PMU) in power system in recent years, modal analysis approaches which are based on real time measurement data from PMU are drawing great interest. Frequency Domain Decomposition (FDD) is able to estimate mode information with several minutes of ambient PMU data. However, there are two steps that make FDD time-consuming. One of them is the Power Spectral Density (PSD) matrix calculation which needs auto-spectrum and cross-spectrum estimation with all data channels. The other step is the application of Singular Value Decomposition (SVD) to PSD matrix. Then Fast Frequency Domain Decomposition (FFDD) was proposed to alleviate the constraints of FDD. It does not require full PSD matrix calculation and also avoids SVD. Usually windowed FFT is utilized in PSD matrix calculation to reduce the power spectrum leakage which leads to inaccurate estimation. In this thesis, the Multi-Taper Method (MTM) which uses Slepian tapers as windowing functions is studied. Proper taper setting of Slepian function in FFDD is proposed when the length of data window is set. The goal is to get an accurate estimation especially when the analysis window and repeat time is small and to improve the responsiveness (under poorly damped conditions) and smoothness (under well damped conditions) of FFDD estimation. Short analysis window is preferred to get high refresh rate of estimation result. The suggested taper setting is tested on the simulated cases from Kundur two area system as well as historical synchrophasor data from real power system.
Metrics
10 File views/ downloads
23 Record Views
Details
- Title
- Synchrophasor based ambient oscillation monitoring using short analysis window
- Creators
- Lu Zhang
- Contributors
- Mani Vaithianathan Venkatasubramanian (Chair)Anjan Bose (Committee Member) - Washington State University, Electrical Engineering and Computer Science, School ofANURAG K. SRIVASTAVA (Committee Member) - Washington State University, Electrical Engineering and Computer Science, School of
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; [Pullman, Washington] :
- Number of pages
- 51
- Identifiers
- 99900525199901842
- Language
- English
- Resource Type
- Thesis