Home // ADVCOMP 2024, The Eighteenth International Conference on Advanced Engineering Computing and Applications in Sciences // View article
Authors:
Xin Xing
Sebastian Ohl
Keywords: Autonomous Driving; Decision-making; Reinforcement Learning; Car-following models; Adaptive Cruise Control; Automatic Emergency Braking; Proximal Policy Optimization; Policy Gradient.
Abstract:
Autonomous driving technology has progressed significantly, necessitating advanced maneuver-based decision-making systems for complex driving environments. Traditional methods often fail in unpredictable real-world scenarios, leading to the adoption of learning-based approaches, such as Deep Learning (DL) and Reinforcement Learning (RL). This paper explores safety-critical car-following models and traffic management, focusing on Adaptive Cruise Control (ACC) and Automatic Emergency Braking (AEB). Traditional mathematical models are limited under extreme conditions, thus this study leverages machine learning to enhance vehicle responsiveness. Specifically, we apply RL to train car-following models. We emphasize policy-based RL methods, including Policy Gradient (PG) and Proximal Policy Optimization (PPO), within a simulated environment. The results demonstrate that PPO converges faster and exhibits fewer errors compared to PG. This study confirms that RL can effectively automate maneuver-based decision-making, highlighting the need for further research in diverse traffic conditions.
Pages: 11 to 16
Copyright: Copyright (c) IARIA, 2024
Publication date: September 29, 2024
Published in: conference
ISSN: 2308-4499
ISBN: 978-1-68558-184-8
Location: Venice, Italy
Dates: from September 29, 2024 to October 3, 2024