Autonomous mobile flock traffic simulation in digital twin mode

Abstract

Traffic congestion in urban areas is the main reason for the long traveling time from one place to another. This happens due to the decisions and reaction times of each driver. Reducing the influence of the driver solution on the control of the vehicle, i.e., increasing the autonomy of the vehicle, can minimize waiting times at traffic light-controlled and uncontrolled intersections. By minimizing the waiting time at the crossroad, the overall traffic intensity can be reduced as well. This research focuses on obtaining information from simulations at specific crossroads for further observations and traffic optimizations, e.g., by implementing machine learning methods. In order to represent the impact of different levels of autonomous vehicles on the autonomous mobile flock traffic, the open-source SUMO (Simulation of Urban MObility) software is used to simulate the traffic in a digital twin mode. The obtained simulation results provide information about the average speed of surrounding vehicles and the number of vehicles over a period of time, with different scenarios reflecting the density ratio of various levels of vehicle autonomy.