| Zeit: | 29. Oktober 2024 |
|---|---|
| Download als iCal: |
|
Prof. Alberto Bemporad
IMT School for Advanced Studies Lucca
Lucca, Italy
Tuesday 2024-10-29 3 p.m.
IST Seminar Room 2.255 - Pfaffenwaldring 9 - Campus Stuttgart-Vaihingen
Abstract
Machine learning has gained immense popularity in various fields, including control, due to its ability to extract mathematical models from data. In my talk, I will present different machine learning techniques that can aid in designing and calibrating model predictive control (MPC) laws. I will present different batch and incremental methods tailored for the identification of nonlinear state-space models, such as recurrent neural networks, possibly under nonsmooth L1 and group-Lasso penalties. I will show evidence that these methods, which are related to the family of quasi-Newton optimization algorithms, tend to outperform popular gradient-descent methods. Moreover, I will discuss the use of incremental learning methods to train disturbance models and achieve offset-free tracking in nonlinear MPC. Finally, I will present global and preference-based optimization techniques that rely on surrogate functions to actively learn the optimal MPC parameters.
Biographical Information
Alberto Bemporad received his Ph.D. in Control Engineering in 1997 from the University of Florence, Italy. In 1996/97 he was with the Center for Robotics and Automation, Department of Systems Science & Mathematics, Washington University, St. Louis. In 1997-1999 he held a postdoctoral position at the Automatic Control Laboratory, ETH Zurich, Switzerland, where he collaborated as a Senior Researcher until 2002. In 1999-2009 he was with the Department of Information Engineering of the University of Siena, Italy, becoming an Associate Professor in 2005. In 2010-2011 he was with the Department of Mechanical and Structural Engineering of the University of Trento, Italy. Since 2011 he has been a Full Professor at the IMT School for Advanced Studies Lucca, Italy, where he served as the Director of the institute from 2012 to 2015. He spent visiting periods at Stanford University, the University of Michigan, and Zhejiang University. In 2011 he co-founded ODYS S.r.l., a company specialized in developing model predictive control systems for industrial production. He is the author or coauthor of various software packages for model predictive control design and implementation, including the Model Predictive Control Toolbox (The Mathworks, Inc.) and the Hybrid Toolbox for MATLAB, and is the co-inventor of 21 patents. He was an Associate Editor of the IEEE Transactions on Automatic Control during 2001-2004 and Chair of the Technical Committee on Hybrid Systems of the IEEE Control Systems Society from 2002 to 2010. He received the IFAC High-Impact Paper Award for the 2011-14 triennial, the IEEE CSS Transition to Practice Award in 2019, the 2021 SAE Environmental Excellence in Transportation Award, the 2024 Beale–Orchard-Hays Prize for Excellence in Computational Mathematical Programming, the Control Engineering Practice 2024 Best Paper Award, and an ERC Advanced Research Grant in 2024. He has been an IEEE Fellow since 2010.