A distributed ADMM method for dynamic economic dispatch over unbalanced directed communication networks

Authors:

Abstract
To address multi-period coupled constraints, nonlinear cost functions, and inaccurate global information acquisition under unbalanced directed communication networks in microgrid dynamic economic dispatch (DED), this paper proposes a distributed alternating direction method of multipliers (ADMM) method for DED. The original problem is decomposed into three subproblems: power update, local feasible-set projection, and dual variable update. In the power update subproblem, Newton-Raphson iteration is introduced to solve the nonlinear optimization problem, while Ratio Consensus is employed to obtain the global average information required for Newton iteration under unbalanced directed communication networks. In the local feasible-set projection step, generation capacity constraints and multi-period ramp-rate constraints are handled by a projection method. Finally, a microgrid case study is carried out to verify the proposed method. Simulation results show that the proposed method can effectively satisfy operational constraints and achieve optimization accuracy close to that of the exact-sum method.

This publication has 7 references indexed in Scilit: