Technical Paper
Deep Reinforcement Learning with Artificial Potential Field for Autonomous Driving Decision-Making in Roundabouts
2024-04-09
2024-01-2871
Roundabouts are one of the most complex traffic environments in urban roads and a key challenge in intelligent driving decision-making. Deep reinforcement learning, as an emerging solution for intelligent driving decisions, has the advantage of avoiding complex algorithm design processes and sustainable iterations. Aiming at the decision-making difficulties in roundabout scenarios, this paper proposes a deep reinforcement learning Soft Actor-Critic algorithm based on the Artificial Potential Field (APF-SAC). Firstly, based on the CARLA simulator and GYM framework, this paper builds a reinforcement learning simulation system for roundabout driving. Secondly, to reduce the exploration difficulty of reinforcement learning, this paper designs global path planning and path smoothing algorithms to generate and optimize paths for guiding the agent.