Connected vehicles and road infrastructure for urban mobility

Abstract

The advancement of transportation led to an increase in vehicle ownership, thereby increasing traffic congestion. Traffic congestion causes several problems, some of which are the increase in air pollution, fuel consumption, collisions, delays and driver’s frustration. To address this problem, a methodology based on Vehicle to Infrastructure (V2I) communication is developed to control the traffic light timings at urban intersections in real time. To implement this, the infrastructure gathers data from V2I communication equipped vehicles, which provides information on the number of vehicles passing through the junction. Based on the number of vehicles, a linear regression model is developed in order to predict the travel time through the junction. The simplex algorithm is used to minimise the error between the predicted travel time and a desired travel time, and generates the ideal green time ratios for this error to be minimised. The chosen junction is tested in different scenarios such as with no traffic lights installed, with traffic lights operating with fixed timings and finally, operating with adaptive traffic light control. When working with adaptive control, the methodology was further validated when the vehicles communicated both their origin and destination, as well as when the vehicles communicated their origin only. Additionally, the system behaviour was further analysed assuming continuously lower rates of vehicles equipped with V2I communication. Results will show that with a rate of equipped vehicles as low as 10 per cent of the total private vehicle fleet, the infrastructure manages to optimise the traffic light timings while minimising the total travel time through the junction.