Embedded Files
Tutorial at SmartGridComm 2021
Abstract
With the increase in solar and wind generation and the deployment of more electric vehicles, the electric power system is going through an unprecedented transformation which render today’s real-time control architecture ineffective. This transition calls for more elaborated real-time control strategies for voltage regulation, line congestion control, and power curtailment. We propose a unified control framework that enables autonomous grid control beyond the state-of-the-art. The key idea is to turn iterative optimization algorithms into robust output-feedback controllers. These controllers, instead of tracking a given reference, drive the system to the optimum of a user-defined optimization problem. Furthermore, these controllers can naturally deal with input and output constraints and are proven to be robust against model mismatch, both in theory and in experiments. In this tutorial, we will present the theoretical foundation of this framework and present the results of a three-year collaboration with the French transmission grid operator, including numerical simulations on a real part of the French subtransmission grid. Furthermore, we will present an experimental demonstration of the controller on a distribution feeder. The attendees will be given access to benchmarks and code to repeat our numerical experiments and compare their own solutions to the proposed approach.
The Theory of Feedback Optimization (SB, 25 min)
The audience will be introduced to the concept of Feedback Optimization, which allows to implement iterative optimization algorithms as robust output-feedback controllers on a physical plant.
Feedback Optimization in Distribution Grids (LO, 25 min)
We will show how feedback optimization can be employed to design a real-time coordinated volt/VAr control scheme for distribution grids with high renewable penetration.
Experimental study: We will illustrate the experimental performance of the proposed controllers, contrast it to standard solutions (e.g. droop), and discuss the lessons learned.
Benchmark 1: A low-voltage distribution test feeder that other researcher can use to evaluate their controllers and compare them to our control algorithm.
Feedback Optimization in Transmission Grids (LO, 25 min)
We will show how the feedback optimization paradigm constitutes a unified approach to multiple aspects of real-time operation of transmission and sub-transmission grids: voltage regulation, line congestion management, and curtailment of renewable generation.
Numerical simulations: We will present and discuss the numerical results of a three-year collaboration with the French TSO Réseau de Transport d’Électricité (RTE).
Benchmark 2: A numerical model based on the French subtransmission grid, where researchers can test different real-time operation strategies in the presence of a substantial renewable generation.
Q&A Session (moderated by SB, 15 min)
As part of the UNICORN project, we will distribute numerical benchmarks to test real-time control strategies for both transmission grids and distribution grids.
Relevant publications
Miguel Picallo, Dominic Liao-McPherson, Saverio Bolognani and Florian Dörfler. Cross-layer design for real-time grid operation: estimation, optimization and power flow. Proc. 22nd Power Systems Computation Conference (PSCC), 2022. [ preprint ]
Miguel Picallo, Lukas Ortmann, Saverio Bolognani and Florian Dörfler. Adaptive real-time grid operation via online feedback optimization with sensitivity estimation. Proc. 22nd Power Systems Computation Conference (PSCC), 2022. [ preprint ]
Adrian Hauswirth, Saverio Bolognani, Gabriela Hug, and Florian Dörfler. Optimization algorithms as robust feedback controllers. arXiv:2103.11329 [math.OC], 2021. [ preprint ]
Miguel Picallo, Saverio Bolognani, and Florian Dörfler. Closing the loop: Dynamic state estimation and feedback optimization of power grids. Electric Power Systems Research, 189:106753, December 2020. Presented at the 21st Power Systems Computation Conference (PSCC). [ DOI | preprint ]
Lukas Ortmann, Adrian Hauswirth, Ivo Caduff, Florian Dörfler, and Saverio Bolognani. Experimental validation of feedback optimization in power distribution grids. Electric Power Systems Research, 189:106782, December 2020. Presented at the 21st Power Systems Computation Conference (PSCC). [ DOI | preprint | video ]
Lukas Ortmann, Alexander Prostejovsky, Kai Heussen, and Saverio Bolognani. Fully distributed peer-to-peer optimal voltage control with minimal model requirements. Electric Power Systems Research, 189:106717, December 2020. Presented at the 21st Power Systems Computation Conference (PSCC). [ DOI | preprint - video ]
Adrian Hauswirth, Alessandro Zanardi, Saverio Bolognani, Florian Dorfler, and Gabriela Hug. Online optimization in closed loop on the power flow manifold. In Proc. 12th IEEE PES PowerTech, July 2017. Best Student Paper Award. [ DOI | preprint ]
Adrian Hauswirth, Saverio Bolognani, Gabriela Hug, and Florian Dörfler. Projected gradient descent on Riemannian manifolds with applications to online power system optimization. In Proc. 54th Annual Allerton Conference on Communication, Control, and Computing, 2016. [ DOI | preprint ]
Speakers
Lukas Ortmann
Ph.D. Student
Automatic Control Laboratory
ETH Zurich
https://control.ee.ethz.ch/people/profile.lukas-ortmann.html
Saverio Bolognani
Senior Researcher
Automatic Control Laboratory
ETH Zurich
Google Sites
Report abuse