Learning Model Predictive Control for Iterative Tasks. A Data-Driven Control Framework

By Ugo Rosolia, Francesco Borrelli. IEEE Transactions on Automatic Control, 2016

366 citations recorded by the source corpus.

A learning model predictive controller for iterative tasks is presented. The controller is reference-free and is able to improve its performance by learning from previous iterations. A safe set and a terminal cost function are used in order to guarantee recursive feasibility and nondecreasing performance at each iteration. This paper presents the control design approach, and shows how to recursively construct terminal set and terminal cost from state and input trajectories of previous iterations. Simulation results show the effectiveness of the proposed control logic.

Subfields: Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/e15369457639ffa1c8ef98860830514c444237c2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

Citations in the corpus

Learning-Based Model Predictive Control: Toward Safe Learning in Control (2020, background): “Rosolia & Borrelli (111) presented constraint and combines it with terminal cost learning, assigning the cost-to-go from previous iterations to the points in the sampled safe set; they showed that the a l s e l y . resulting overall cost of the trajectory of an episode is nonincreasing over the…”
A Review of Safe Reinforcement Learning: Methods, Theory and Applications (2022, background): “control [17, 132, 139, 183, 204, 231, 258], adaptive control [89, 115, 199, 238], robust control [26, 73, 112, 282], lyapunov function [56, 57] for RL stability and safety.”
Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning (2021, methodology): “Further performance improvement is achieved by combining robust adaptive MPC with iterative learning (57), which updates the terminal constraint set Xterm in Equation 12d and the terminal cost lH in Equation 12a after every full iteration using the closed-loop state trajectory and cost (58).”
A Tour of Reinforcement Learning: The View from Continuous Control (2018, background): “So what if we learn to iteratively improve the Q-function while running RHC? This idea has been explored in a project by Rosolia, Carvalho, and Borrelli [61].”