Distributed Optimal Economic Dispatch for Smart Grids Based on Multi-Agent System

Abstract

In this works, the economic dispatch problem for smart grids is simplified by the $\varepsilon$–exact penalty functions. A novel and systematic distributed optimization algorithm is introduced based on the multi-agent system via the algebraic graph theory. In order to minimize the operation cost in the entire grid, cooperative control of the power outputs of all power generators in the smart grid will be presented by using of the novel optimization algorithm. Furthermore, through the iteration of the penalty parameters and smooth parameters, the convergence of the algorithm is proved. The solution of the penalty optimization problem coincides with the optimal solution of the economic dispatch problem. A numerical example testifies the good performance of the new algorithm.