Probabilistic Model Predictive Safety Certification for Learning-Based Control

• Published in: IEEE transactions on automatic control Vol. 67; no. 1; pp. 176 - 188
• Main Authors: Wabersich, Henk J., Hewing, Lukas, Carron, Andrea, Zeilinger, Melanie N.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.01.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Bayesian analysis; Certificates; Certification; Learning; Optimization; Prediction algorithms; Predictive control; Predictive models; Probabilistic logic; Probabilistic models; Probability theory; reinforcement learning (RL); Safety; Safety critical; Statistical inference; stochastic systems; Tubes; Uncertainty

Reinforcement learning (RL) methods have demonstrated their efficiency in simulation. However, many of the applications for which RL offers great potential, such as autonomous driving, are also safety critical and require a certified closed-loop behavior in order to meet the safety specifications in...