Abstract
We propose a novel data-driven method to accelerate the convergence ofAlternating Direction Method of Multipliers (ADMM) for solving distributed DCoptimal power flow (DC-OPF) where lines are shared between independent networkpartitions. Using previous observations of ADMM trajectories for a given systemunder varying load, the method trains a recurrent neural network (RNN) topredict the converged values of dual and consensus variables. Given a newrealization of system load, a small number of initial ADMM iterations is takenas input to infer the converged values and directly inject them into theiteration. We empirically demonstrate that the online injection of these valuesinto the ADMM iteration accelerates convergence by a significant factor forpartitioned 14-, 118- and 2848-bus test systems under differing load scenarios.The proposed method has several advantages: it maintains the security ofprivate decision variables inherent in consensus ADMM; inference is fast and somay be used in online settings; RNN-generated predictions can dramaticallyimprove time to convergence but, by construction, can never result ininfeasible ADMM subproblems; it can be easily integrated into existing softwareimplementations. While we focus on the ADMM formulation of distributed DC-OPFin this paper, the ideas presented are naturally extended to other distributedoptimization problems.