Einladung zum Vortrag im Kolloquium Technische Kybernetik

  Predictive control of large-scale systems using distributed optimization

Prof. Dr. Tamás Keviczky
Delft Center for Systems and Control
Delft University of Technology
Netherlands

Tuesday, 31. May 2011, 4:00 p.m.
IST-Seminar-Room 3.243 - Pfaffenwaldring 9 - Campus Stuttgart-Vaihingen

Abstract

Control of large-scale industrial processes and infrastructure systems requires the coordination of interacting subsystems while striving for optimal operation that enforces critical operational constraints. Due to its ability to handle important process constraints explicitly, Model Predictive Control (MPC) has become the method of choice when designing control systems for such applications. However, repeatedly solving the underlying finite-time optimal control problems online may become prohibitive for large-scale systems due to computational or communication constraints.
Recent efforts have been focusing on how to decompose the underlying optimization problem in order to arrive at a distributed or hierarchical control system that can be implemented under the prescribed computational and communication limitations. In the first part of this talk, I will present a subgradient method for solving coupled optimization problems in a distributed way given restrictions on the communication topology. The resulting iterative solution procedure relies on local subgradient updates in combination with a consensus process. In the second part, a hierarchical MPC approach will be described for large-scale systems based on dual decomposition. The proposed scheme allows coupling in both dynamics and constraints between the subsystems and generates a primal feasible solution within a finite number of iterations, using primal averaging and a constraint tightening approach.

Biographical Information

Tamás Keviczky is an Assistant Professor at the Delft Center for Systems and Control, Delft University of Technology in Delft, The Netherlands. He received the MSc degree in Electrical Engineering in 2001 from the Budapest University of Technology and Economics, and the PhD degree in 2005 from the Control Science and Dynamical Systems Center at the University of Minnesota, Minneapolis. Between 2005 and 2007 he has been a Research Intern at Honeywell Laboratories in Minneapolis, and a Postdoctoral Scholar at the California Institute of Technology, in the Control and Dynamical Systems Department. He was a co-recipient of the AACC O. Hugo Schuck paper award for practice in 2005. His current research interests include distributed optimization and model predictive control, distributed control and estimation in multi-agent robotic systems and large-scale industrial/infrastructure networks.