An Integrated Framework for Distributed Model Predictive Control of Large-Scale Power Networks

Abstract

This work proposes a novel integrated framework in order to model, simulate, and optimize potentially large-scale and complex networks, fusing an extended network modeling methodology and distributed model predictive control structures based on Lagrange multipliers. In concrete, a revised network modeling formulation is utilized, allowing extended networking capabilities with a novel, generic, and compact way of organizing network topology information. This optimization framework offers a number of very convenient features: distribution of the overall network control effort among the local agents; no overall network knowledge required; the introduction/elimination of a node only affects to its neighbors. Thus, this integrated framework offers a powerful and easy-to-use mathematical tool in order to analyze and manage a wide variety of today's large-scale network. Specifically, this methodology is applied here to a power network to show the benefit of the work developed.