Performance-oriented model learning for data-driven MPC design

Abstract

Model Predictive Control (MPC) is an enabling technology in applications requiring controlling physical processes in an optimized way under constraints on inputs and outputs. However, in MPC closed-loop performance is pushed to the limits only if the plant under control is accurately modeled; otherwise, robust architectures need to be employed, at the price of reduced performance due to worst-case conservative assumptions. In this paper, instead of adapting the controller to handle uncertainty, we adapt the learning procedure so that the prediction model is selected to provide the best closed-loop performance. More specifically, we apply for the first time the above "identification for control" rationale to hierarchical MPC using data-driven methods and Bayesian optimization.

Accepted for publication in the IEEE Control Systems Letters (L-CSS)

References in corpus (1)

• Direct data-driven control of constrained linear parameter-varying systems: A hierarchical approach

Cited by in corpus (7)

• On Recurrent Neural Networks for learning-based control: recent results and ideas for future developments
• Performance-based Trajectory Optimization for Path Following Control Using Bayesian Optimization
• Robust constrained nonlinear Model Predictive Control with Gated Recurrent Unit model -- Extended version
• From system models to class models: An in-context learning paradigm
• Learning safety in model-based Reinforcement Learning using MPC and Gaussian Processes
• Maximum Likelihood Identification of Linear Models with Integrating Disturbances for Offset-Free Control
• A predictive modular approach to constraint satisfaction under uncertainty -- with application to glycosylation in continuous monoclonal antibody biosimilar production