Current Research Topics in Distributed Optimization of Smart Grids

Speaker:    Mr. Philipp Sauerteig

Time:        15:00-16:00, June 10

Location:    SIST 1C-201

Host:       Prof. Boris Houska

Abstract:

Electrical networks go through a profound transition. In the past, energy was produced and stored centrally organized by the grid operator only. Nowadays, more and more renewable energy sources such as photovoltaic panels and wind turbines are installed in the grid. The additional production results in the need to store superfluous energy temporarily in domestic energy storage units (batteries). Hence, energy is both produced and stored in a distributed way. This fundamental change in the system leads to optimization problems that cannot be solved centralized. The first part of this talk provides an introduction to distributed optimization of smart grids in a Model Predictive Control (MPC) framework. Based on a simple model describing the system dynamics of a single prosumer (producer + consumer) a state-of-the-art Alternating Direction Method of Multipliers (ADMM) is used to solve a convex optimization problem arising in modern smart grids. The second part of the talk is dedicated to current research topics. First, an extension of the above mentioned problem is given in order to apply a recently developed Augmented Lagrangian based Alternating Direction Inexact Newton method (ALADIN). Moreover, the problem is considered in a multi objective optimization framework. In this context the concept of Pareto optimality is used to find optimal control strategies. Finally, the additional possibility to exchange energy within coupled micro grids is used to improve the overall performance. To this end a bi level optimization problem is investigated. Negotiation between the two layers comes along with a high communication effort. As a remedy a strategy is proposed, where the optimization step on the lower level is replaced by surrogate models.

Bio:

Philipp Sauerteig received both his B.S. (2016) and his M.S. degree (2017) in mathematics from Technical University Ilmenau, Germany. He started his Ph.D. in 2018. Currently, Mr. Sauerteig works as a research assistant at his alma mater funded by the federal ministry of education and research on the project “CONSENS: Consistent Optimization and Stabilization of Electrical Networked Systems”. His research concerns distributed optimization, model predictive control, multi criteria optimization and surrogate models.

SIST-Seminar 18170