Optimal scheduling of a reconfigurable active distribution network with multiple autonomous microgrids

Abstract

In this paper, a decentralized framework for optimal scheduling of a reconfigurable Active Distribution Network (ADN) is proposed that preserves the autonomy and information privacy of all the agents, i.e., the distribution system and microgrid's operators. In the proposed framework, the scheduling problem is decomposed to a certain number of Local Scheduling Models (LSMs), in such a way that each agent can independently schedule its network. In the proposed LSM, all technical constraints such as AC load flow equations, and radiality constraints are respected. Also, a Multi-Level Analytical Target Cascading (MLATC)-based method is proposed to coordinate the scheduling results of the LSMs. In this framework, by introducing new auxiliary variables, the shared variables that couple the LSMs are limited to only three types. Using the proposed framework, the optimal configuration of the ADN and the operational scheduling of agents are simultaneously calculated. Since all technical constraints are modeled in the proposed LSM, the results are both economic and feasible. To assess the optimality and effectiveness of the proposed method, several simulations are carried on the modified IEEE 33-bus and IEEE 123-bus distribution test systems. Moreover, the results are compared with the centralized approach to validate the optimality of the solution.