Print Email Facebook Twitter Near optimal control with reachability and safety guarantees Title Near optimal control with reachability and safety guarantees Author Verdier, C.F. (TU Delft Team Tamas Keviczky) Babuska, R. (TU Delft Learning & Autonomous Control) Shyrokau, B. (TU Delft Intelligent Vehicles) Mazo, M. (TU Delft Team Tamas Keviczky) Date 2019 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. Subject nonlinear optimal controlreinforcement learningvalue iterationvehicle safetyverification To reference this document use: http://resolver.tudelft.nl/uuid:34abfeb3-f3c7-4216-a3a8-f667a8ed592b DOI https://doi.org/10.1016/j.ifacol.2019.09.146 ISSN 1474-6670 Source IFAC-PapersOnLine, 52 (11), 230-235 Event 5th IFAC Conference on Intelligent Control and Automation Sciences, ICONS 2019, 2019-08-21 → 2019-08-23, Belfast, United Kingdom Part of collection Institutional Repository Document type journal article Rights © 2019 C.F. Verdier, R. Babuska, B. Shyrokau, M. Mazo Files PDF 1_s2.0_S2405896319307761_main.pdf 625.37 KB Close viewer /islandora/object/uuid:34abfeb3-f3c7-4216-a3a8-f667a8ed592b/datastream/OBJ/view