Towards Real Time Control of Oscillations in Electric Power Systems

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This proposal addresses the problem of effective security control of extensive, heavily loaded power systems, focusing on the study of mechanisms and precursors of widespread oscillations, and on the problems of selecting effective controls in the face of limited data. Its ultimate objective is to provide system operators with abilities to detect the imminence of potentially dangerous oscillations and to take preventive measures to avoid danger to equipment or the system. It is confined to consideration of spontaneous oscillations caused by gradual changes in system conditions, and to oscillations due to disturbances. "Forced" oscillations resulting from incomplete islanding or pulsating loads are not considered. The proposal draws heavily on the recently defined concept of "strong resonance." It is proposed that two lightly damped oscillatory modes interact near a resonance. "If linearization is not diagonizable, the resonance is called "strong;" if it is diagonizable, it is called "weak." An eigenvalue that changes in frequency before the resonance can change in damping after the resonance and become oscillatory unstable as the damping changes through zero." Thus, modal resonance is considered as a possible precursor of power system oscillations.