ARC@ORU Combining Reinforcement Learning and Model Predictive Control - How to take optimal decisions with wrong models?

Venue: ARC, Örebro University

Remote participation: Join by this Zoom link

Registration: Please register by the form below.

About the seminar

Reinforcement Learning (RL) can be used to tune Model Predictive Control (MPC) schemes towards closed-loop optimality in the real world. That tuning is fundamentally different than using classical fitting approaches to adjust the model underlying MPC. These ideas are supported by a rich theory on how MPC is a universal approximation for Markov Decision Processes, provided that it is viewed in the right way.

That theory proposes a bridge between the open-loop predictions of MPC and its closed-loop implementation. It also explains why using classical model fitting is sometimes good enough for achieving near-optimal closed-loop performance with MPC, and why it often falls short of achieving that. It finally opens a theoretical pathway on using (imperfect) simulation models to build optimal policies for the real-world, and on closing the “sim2real gap”.

Speaker

Sebastien Nicolas Gros is a Professor at the Norwegian University of Science and Technology (NTNU).

Sebastiens current research concerns safe Reinforcement Learning and Data-Driven MPC. He currently has 6 PhD students working on connected topics. This work is strongly connected to the industry, with large industrial partners such as Equinor, DNV, Konsberg, but also smaller ones focusing on energy.

Sebastien also co-supervise several PhD projects related to the Artificial Pancreas problem, and multi-rotor wind turbine control. Finally, he supervise and co-supervise several industrial PhD projects based at Volvo (Sweden), focusing on powertrain control, 2nd life of batteries, and traffic management. 

In this talk, we will provide an overview of these recent advances.