Chance-Constrained Control with Lexicographic Deep Reinforcement Learning

no code implementations • 19 Oct 2020 • Alessandro Giuseppi, Antonio Pietrabissa

Standard DeepRL approaches require i) the constraints to be included as additional weighted terms in the cost function, in a multi-objective fashion, and ii) the tuning of the introduced weights during the training phase of the Deep Neural Network (DNN) according to the probability thresholds.

reinforcement-learning Reinforcement Learning (RL)