Distributed Optimization Methods for Wide-Area Damping Control of Power System Oscillations

This paper presents a distributed optimization algorithm to address the problem of wide-area damping control of large-scale electric power systems using Synchrophasors. Our approach consists of a three-step strategy First, Synchrophasors from selected nodes in a power network are used to identify offline dynamic models of the dominant areas of the network. Thereafter, a linear controller is designed for this reduced-order model to shape the inter-machine oscillation dynamics. Finally, algorithms are developed to invert this design to realistic local controllers in each area by optimizing the controller parameters until their interarea response matches the closed-loop inter-machine response achieved in the second step. A model reference control design following this three-step strategy was recently proposed in [1] using a centralized controller. Our results in this paper extend that design by posing the problem purely from a perspective of distributed optimization.