Publication
Hybrid Deep Reinforcement Learning and Planning for Safe and Comfortable Automated Driving
Dikshant Gupta; Matthias Klusch
In: Intelligent Vehicles. IEEE Intelligent Vehicles Symposium (IV-2023), IEEE, 2023.
Abstract
We present a novel hybrid learning method, named HyLEAR, for solving the collision-free navigation problem for self-driving cars in POMDPs. HyLEAR leverages interposed learning to embed knowledge of a hybrid planner into a deep reinforcement learner to faster determine safe and comfortable driving policies of the car. In particular, the hybrid planner combines pedestrian path prediction and risk-aware path planning with driving-behavior rule-based reasoning such that the determined safe trajectories also take into account, whenever possible, the ride comfort and a given set of driving behavior rules. Our experimental performance analysis over the CARLA-CTS benchmark of critical traffic scenarios revealed that HyLEAR can significantly outperform the selected baselines in terms of safety and ride comfort.