Hybrid Deep Reinforcement Learning and Planning for Safe and Comfortable Automated Driving

TL;DR

This paper introduces HyLEAR, a hybrid deep reinforcement learning method that integrates planning and learning to improve safety and comfort in autonomous driving within complex traffic scenarios.

Contribution

HyLEAR is a novel hybrid approach combining planning and deep reinforcement learning for collision-free navigation in POMDPs, emphasizing safety and ride comfort.

Findings

HyLEAR outperforms baselines in safety metrics.

HyLEAR enhances ride comfort in critical scenarios.

Experimental results validate the effectiveness of the hybrid approach.

Abstract

We present a novel hybrid learning method, 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. In particular, the hybrid planner combines pedestrian path prediction and risk-aware path planning with driving-behavior rule-based reasoning such that the driving policies also take into account, whenever possible, the ride comfort and a given set of driving-behavior rules. Our experimental performance analysis over the CARLA-CTS1 benchmark of critical traffic scenarios revealed that HyLEAR can significantly outperform the selected baselines in terms of safety and ride comfort.