Automated Lane Change Decision Making using Deep Reinforcement Learning in Dynamic and Uncertain Highway Environment

Automated Lane Change Decision Making using Deep Reinforcement Learning in Dynamic and Uncertain Highway Environment Ali Alizadeh 1, Majid Moghadam 2, Y unus Bicer 3, Nazim Kemal Ure 4, Ugur Y avas 5 and Can Kurtulus 5 Abstract -- Autonomous lane changing is a critical feature for advanced autonomous driving systems, that involves several challenges such as uncertainty in other driver's behaviors and the tradeoff between safety and agility. In this work, we develop a novel simulation environment that emulates these challenges and train a deep reinforcement learning agent that yields consistent performance in a variety of dynamic and uncertain traffic scenarios. Results show that the proposed data-driven approach performs significantly better in noisy environments compared to methods that rely solely on heuristics. I NTRODUCTION Advanced Driving Assistance Systems (ADAS) are developed to increase traffic safety by reducing the impact of human errors. The evolution of various levels of driving autonomy has seen a significant speedup in last years aiming to enhance comfort, safety, and driving experience. For a long time, with a limited amount of technological resources, automotive stakeholders were focusing on steady-state maneuvers to achieve driving autonomy.