AC4MPC: Actor-Critic Reinforcement Learning for Nonlinear Model Predictive Control

no code implementations • 6 Jun 2024 • Rudolf Reiter, Andrea Ghezzi, Katrin Baumgärtner, Jasper Hoffmann, Robert D. McAllister, Moritz Diehl

The \ac{RL} critic is used as an approximation of the optimal value function, and an actor roll-out provides an initial guess for primal variables of the \ac{MPC}.

Model Predictive Control reinforcement-learning

Imitation Learning from Nonlinear MPC via the Exact Q-Loss and its Gauss-Newton Approximation

no code implementations • 3 Apr 2023 • Andrea Ghezzi, Jasper Hoffman, Jonathan Frey, Joschka Boedecker, Moritz Diehl

This work presents a novel loss function for learning nonlinear Model Predictive Control policies via Imitation Learning.

Imitation Learning Model Predictive Control