Prediction of Generators' Participation Factors and Oscillation Types for Dominant Oscillatory Modes of Power System

Abstract

Small-signal rotor angle stability, also known as small-signal stability, is an important issue for power system operation. This stability is associated with inter-area, local, control, and torsional oscillations. To cope with it, the type and main drivers of the oscillations should be determined. In this way, the system operators can plan more effective preventive and corrective actions, e.g., to transfer power generation from more vulnerable units to less vulnerable ones. However, determining oscillation modes and their participation factors by means of conventional methods (e.g., time-domain simulation and modal analysis) is a challenging and time-consuming task, which is not appropriate for on-line environments, such as dispatching centers of power systems. In this article, a new viewpoint for this problem is proposed through modeling it as a forecast process by which the participation factors of generators for dominant modes as well as the oscillation types are predicted. A new prediction strategy, composed of an information-theoretic feature selection, a probabilistic neural network, and a line search procedure, is also presented to implement the forecast process. The effectiveness of the proposed approach for small-signal stability evaluation is extensively illustrated on the IEEE New England test system.