On the effects of distributed control area design for the stabilization of cyber-enabled smart grids

Abstract

We study the effect of control area design on the performance of distributed control. Specifically, we consider distributed implementations of parametric feedback linearization (PFL) control for efficient transient stability after the occurrence of a power system disturbance. We employ hierarchical spectral clustering and k-mean spectral clustering techniques to design control areas with high physical coupling within the power system. We address three distributed control scenarios: (i) distributed control applied to all generators of a control area, (ii) distributed control applied only to the largest inertia generator within a control area, and (iii) hierarchical distributed control where all generators apply distributed control and lead generators within a control area have centralized control. We investigate the effect of area clustering outcomes and compare the performance of the three control approaches for various power system faults.