Effect of Number of Lanes on Traffic Characteristics of Reinforcement Learning Based Autonomous Driving

Traffic characteristics such as signalized intersections and high vehicle density combined with low speeds account for 12-55% increase in commute time. Increasing the number of lanes highway infrastructure can help capacity consequently reduce associated delay. However, due to human-related features driver behavior interaction well induced demand, it is recommended limit four. Recently, view rising discussion regarding deployment Autonomous Vehicles (AVs), important study effect on traffic respect AVs. This will provide insights into full potential AVs terms reducing time crucial design future road networks. Therefore, this study, RL-based AV frameworks are developed investigate characteristics. Specifically, we travel speed, collision, driving right-most lane. Results two well-known simulated highways ranging from 2 8 increasing count respectively show improved increases. improvement depends RL algorithm employed.