Near optimal control with reachability and safety guarantees

Journal article (2019)

Authors

C.F. Verdier Team Tamas Keviczky - Mechanical, Maritime and Materials Engineering
R. Babuska Learning & Autonomous Control - Mechanical, Maritime and Materials Engineering
B. Shyrokau Intelligent Vehicles - Mechanical, Maritime and Materials Engineering
M. Mazo Team Tamas Keviczky - Mechanical, Maritime and Materials Engineering
Research Group
Team Tamas Keviczky (Mechanical, Maritime and Materials Engineering) (TU Delft)
Copyright: © 2019 C.F. Verdier, R. Babuska, B. Shyrokau, M. Mazo
DOI
help_outline
https://doi.org/10.1016/j.ifacol.2019.09.146
Reinforcement learning Verification Vehicle safety Value iteration Nonlinear optimal control
More Info
expand_more

Published Date

2019

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Mechanical, Maritime and Materials Engineering

Department

Delft Center for Systems and Control

Research Group

Team Tamas Keviczky

Abstract

Control systems designed via learning methods, aiming at quasi-optimal solutions, typically lack stability and performance guarantees. We propose a method to construct a near-optimal control law by means of model-based reinforcement learning and subsequently verifying the reachability and safety of the closed-loop control system through an automatically synthesized Lyapunov barrier function. We demonstrate the method on the control of an anti-lock braking system. Here the optimal control synthesis is used to minimize the braking distance, whereas we use verification to show guaranteed convergence to standstill and formally bound the braking distance.