Fully Distributed Dynamic Optimal Power Flow of Active Distribution Networks with Multi-microgrids Based on Cutting Plane Consensus Method and Ward Equivalent

Abstract

With distributed energy resources widely integreted to an distribution network (DN) in multiple microgrid (MG) manner, the distributed dynamic optimal power flow (D-DOPF) becomes a major concern because DN and MGs may belong to different owners. This paper proposes a fully D-DOPF algorithm based on the cutting plane consensus (CPC) method and Ward equivalent. First, DN and MGs are decoupled by means of Ward equivalent, i.e., at optimizing each MG, other MGs and DN are replaced by their Ward equivalent circuits respectively, while at optimizing DN, MGs are replaced by their Ward equivalent circuits respectively. Then, high-precision linearization of nonlinear power flow equations is used to express node voltages approximately by a linear function of node injection powers respectively for each MG and DN. Hence, the D-DOPF model is built as a distributed quadratic programming (D-QP) problem, taking DN and MGs as independent agents. Finally, CPC is applied to solve D-DOPF. For each agent, a master problem is constructed to approximate the original problem, and only the cutting plane constraints are transferred between agents. This method does not need an upper level coordinator and also guarantees that each agent has good confidentiality.