Accelerated Methods for the SOCP-relaxed Component-based Distributed Optimal Power Flow

TL;DR

This paper introduces a novel accelerated subgradient method with adaptive penalty parameters to significantly speed up distributed optimal power flow solutions based on SOCP relaxation, achieving up to 87% faster convergence on large-scale systems.

Contribution

It is the first to combine adaptive penalty parameters with accelerated subgradient methods for distributed OPF using SOCP relaxation.

Findings

Achieved up to 87% reduction in convergence time on large test systems.

Demonstrated effectiveness on real-world systems with over 9000 buses.

Provided a new scheme that outperforms existing methods in speed.

Abstract

In light of the increased focus on distributed methods, this paper proposes two accelerated subgradient methods and an adaptive penalty parameter scheme to speed-up the convergence of ADMM on the component-based dual decomposition of the second-order cone programming (SOCP) relaxation of the OPF. This work is the first to apply an adaptive penalty parameter method along with an accelerated subgradient method together in one scheme for distributed OPF. This accelerated scheme is demonstrated to reach substantial speed-ups, as high as 87%, on real-world test systems with more than 9000 buses, as well as on other difficult test cases.