Analyzing the Effects of Automated Vehicles on Urban Roadway Mobility Using Microscopic Simulation
In terms of safety and efficiency, automated vehicles (AVs) are anticipated to offer several advantages to the individual, the transportation system, and society. AVs can be designed to follow the traffic rules and can travel with a shorter headway resulting in improved traffic performance. The chief aim of this study was to investigate the effects of AVs on urban speed, travel time, waiting time, and level of service (LOS) using the microsimulation software SUMO. The prerequisites of a microsimulation tool are car-following and lane-changing models. The car-following model was adopted from previous studies and the lane change model was calibrated using machine learning techniques: second order multivariate regression minimization using Particle Swarm Optimization (PSO). Next, to comprehend how mixed traffic flow conditions might affect the urban traffic performance, a real-world urban corridor, E Commerce ST located in downtown San Antonio, Texas was studied for the peak hour (PH) and double the PH flow for twelve different scenarios. The latter was done to understand the capability of the current infrastructure to hold future demands. The results showed that AVs have the potential to increase urban mobility and reduce the burden of increase in traffic volume. In terms of average speed, travel time, and waiting time, the benefits of adopting AVs were more pronounced under increased flows. In terms of LOS, out of twelve scenarios all but one scenario (scenario 1 of Eastbound (EB) traffic when demand was doubled), degraded from LOS C to LOS D.