Degree Type

Dissertation

Date of Award

2001

Degree Name

Doctor of Philosophy

Department

Electrical and Computer Engineering

First Advisor

Vijay Vittal

Second Advisor

Mustafa Khammash

Abstract

Modern power systems are operated more stressed than ever because of the advent of deregulation and competition. One of the important issues in the design of controllers for a stressed system is to evaluate the stability of the controlled system over a range of operating conditions.;The conventional controllers are designed to make the system stable under certain conditions of operation. The time consuming time domain simulation is then used to evaluate the controllers for a few selected operating conditions around which the controllers are designed. Such a design and evaluation procedure cannot guarantee robustness of the controller over the whole range of operating conditions.;In this dissertation, practical algorithms to perform robustness analysis based on two tools, structured singular value and the nu gap metric, are investigated. The power system stabilizer is used as the controller and small signal stability is of interest.;The robustness problem in the SSV framework is set up for the multimachine power system. In this formulation, an improved uncertainty characterization has been used to capture the effect of parameter variations in terms of the varying elements of the linearized system matries, which are derived from the component differential equations and the network algebraic equations separately. SVD decomposition is used to reduce the size of the problem. Based on the resulting framework, a branch and bound scheme is proposed to intelligently select frequency intervals on which the frequency sweep test can be performed further to find the peak of mu. Instead of blindly choosing frequency intervals to sweep, which could ignore important frequency points on the mu plots, this scheme provides searching under guidance. The analysis procedure accurately predicts the range of stable operating conditions which are verified by repeated eigenvalue analysis.;Fir the robustness in terms of nu gap metric, we set up the feedback configuration for multimachine power system. The frequency response of the nu gap metric is plotted and its relationship with that of the stability margin is used to determine the stability of the perturbed systems. A weighted nu gap metric is defined and its frequency domain interpretation is explored to further reduce the conservatism of the results.;Finally, a feedback configuration is carefully developed to carry out the McFarlane and Glover Hinfinity loop shaping design procedure. The effect of the damping controller on improving system dynamic performance is also examined.;Comparisons are made between the two major analysis tools via the results on the same test systems with the same scenarios.

DOI

https://doi.org/10.31274/rtd-180813-13207

Publisher

Digital Repository @ Iowa State University, http://lib.dr.iastate.edu

Copyright Owner

Chuanjiang Zhu

Language

en

Proquest ID

AAI3016760

File Format

application/pdf

File Size

124 pages

Share

COinS