Organizers
Sasa V. Rakovic, University of Texas, Austin William S. Levine, University of Maryland, College Park Behcet Acikmese, University of Washington, Seattle Ilya V. Kolmanovsky, University of Michigan, Ann Arbor
Model Predictive Control
Model Predictive Control (MPC) has established itself as a dominant advanced control technology across many industries due to its exceptional ability to explicitly account for control objectives, directly handle static and dynamic constraints and systematically optimize performance. MPC provides, via an iterative open loop optimal control implemented by repetitive online optimization, a feedback control that meets design specifications and maximally utilizes system capabilities. Not surprisingly, MPC has also become a highly vibrant and interdisciplinary branch of mathematical control theory that fuses and synergistically treats control–theoretic issues (such as stability, performance, robustness, etc.) for constrained systems with the optimization theory and numerical computations. Important areas in MPC that have recently seen significant theoretical and implementational progress include robust and stochastic MPC as well as efficient computations for MPC via convex and reliable real-time optimization.
Overview
The workshop introduces its audience to the theory, design and applications of model predictive control under uncertainty. The workshop provides conceptual and practical principles governing rigorous and computationally effective methods for design of MPC under set-membership and probabilistic uncertainty. The theoretical fundamentals are carefully introduced and studied within the frameworks of robust and stochastic MPC. The technical foundations are complemented with a study of related design and practical aspects as well as with an overview of effective computations based on convex and reliable real-time optimization. Thus, the workshop provides a concise and unified exposure to MPC under uncertainty.
Themes
William S. Levine
Conventional MPC This part introduces MPC, provides its formulation, discusses its algorithmic implementation and summarizes its fundamental system theoretic properties. It also discusses effects of uncertainty in MPC in terms of types and models as well as interplay of the uncertainty with predictions, constraints and cost. This part also comments on inherent robustness of MPC and provides a base for design methods that take the uncertainty into account more directly, i.e. robust and stochastic MPC.
Sasa V. Rakovic
Robust MPC This part focuses on exact robust MPC and, motivated by its computational intractability, it highlights importance of careful use of parametrized control policies in order to computationally simplify the exact robust MPC whilst preserving as many of its strong structural properties as possible. A particular emphasis is given to tube MPC framework that addresses effectively the fundamental challenge of reaching a meaningful compromise between the quality of guaranteed structural properties and the associated computational complexity. This part also builds upon generic introduction to tube MPC by providing an overview (in terms of theory and design aspects for) of basic and advanced tube MPC design methods, namely rigid, homothetic and parameterized tube MPC synthesis.
Ilya V. Kolmanovsky
Stochastic MPC In this part of the workshop, variants of MPC frequently referred to as stochastic MPC (SMPC) are considered. The presentation of SMPC will begin with a more in-depth discussion of modeling/representation of stochastic uncertainties and probabilistic handling of constraints. Then, existing approaches to SMPC for linear systems are discussed. The presentation continues with extension to the case of SMPC for general nonlinear systems. Mechanisms for guaranteeing recursive feasibility and stochastic stability properties in SMPC problems are also described. Finally, special problems such as of dual (i.e., combined identification/estimation and tracking) control and drift counteraction/optimal stopping are introduced, and their treatment within MPC, are covered as motivated by practical applications.
Behcet Acikmese
Convexification for MPC Under Uncertainty with Reliable Online Computations This part presents methods of convexification and real-time convex optimization for robust and stochastic MPC problems. It begins with an introduction of recent analytical results enabling the formulation of a class of MPC problems within convex optimization framework, as well as presenting linear matrix inequality based methods for handling deterministic and probabilistic disturbances in MPC with particular emphasis on the model uncertainties described via incremental quadratic inequalities. The presentation is concluded with a summary of recent advances in real-time optimization and convex optimization. A particular emphasis is placed on the development of custom Interior Point Method algorithms and methods for customization and autocoding that lead to real-time implementable software.
Overview of Applications and Closing Open Discussion (All organizers) Applications of MPC under uncertainty have been reported in many domains, including finance, building control, electric power grid, chemical process industry, and automotive and aerospace systems. The applications to automotive and aerospace systems are of special interest as these systems operate in uncertain environment, have fast dynamics and very limited onboard computing power. Consequently, these application are in the focus of the workshop. A closing open discussion aims at assessing the current state of affairs in, and identifying relevant future research directions in terms of theory and applications.
Primary Objectives
The major goals of the workshop are to provide a comprehensive tutorial of both fundamental and advanced aspects of MPC under uncertainty and present a unified treatment of its conceptual and practical aspects. More specifically, the workshop delivers a compact understanding of MPC as well as comprehensive theoretical foundation underpinning MPC under both the set–membership and stochastic uncertainty (a.k.a. robust and stochastic MPC). The workshop also covers the numerical implementation of conventional, robust and stochastic MPC design methods via advanced and reliable real–time optimization techniques. Thus, the workshop synergistically fuses theoretical, computational and applications–driven aspects of MPC under uncertainty and, consequently, furnishes a unique blend of advanced control synthesis and analysis methods. An equally important aim of the workshop is to highlight the importance of MPC under uncertainty and disseminate the knowledge of advanced robust and stochastic MPC as powerful design methods with high potential to successfully tackle real–life problems across a wide range of traditional and emerging industrial applications.
Attendee Benefits
The workshop is designed carefully and flexibly in order to be accessible to a broad range of researchers and engineers within both academia and industry. It is entirely presented by the four organizers in a coherent and effective manner. The workshop provides, for junior researchers and students, a comprehensive exposure to advanced theory and design of MPC for constrained systems subject to uncertainty. On the other hand, the workshop delivers a systematic framework for senior researchers and engineers working on real-life industrial problems where constraints and uncertainty play a key role. A closing open panel discussion assesses the current state of affairs in, and identifies relevant future research directions for this field in terms of theory and applications. This one full day event aims to stimulate the creation of a specialized network of researchers focused on further advances in this highly important research field.