Distributed Optimization for Energy Grids: A Tutorial on ADMM and ALADIN

Lukas Lanza,Timm Faulwasser,Karl Worthmann

2024-04-05

Abstract:The ongoing transition towards energy and power systems dominated by a large number of renewable power injections to the distribution grid poses substantial challenges for system operation, coordination, and control. Optimization-based methods for coordination and control are of substantial research interest in this context. Hence, this chapter provides a tutorial introduction of distributed optimization algorithms for energy systems with a large share of renewables. Specifically, we focus on the Alternating Direction Method of Multipliers (ADMM) and on the Augmented Lagrangian Alternating Direction Inexact Newton (ALADIN) method as both algorithms are frequently considered for coordination and control of power and energy systems. We discuss the application of ALADIN and ADMM to AC optimal power flow problems and to energy management problems. Moreover, we give an outlook on open problems.

Optimization and Control

What problem does this paper attempt to address?

This paper discusses the challenges of operation, coordination, and control in power systems with the increasing use of renewable energy and distributed resources. The focus of the study is on the Alternating Direction Method of Multipliers (ADMM) and the Augmented Lagrangian Alternating Direction Inexact Newton (ALADIN) methods, which are commonly used for coordination and control in power systems. The paper introduces the application of these algorithms in AC optimal power flow problems and energy management problems, and highlights future research directions, including open problems.