Reducing Traffic Congestion in Urban Areas via Real-Time Re-Routing: A Simulation Study

Abstract

Traffic congestion problems of urban road networks are having a strong impact on economy, due to losses from accidents and delays, and to public health. The recent progress in connected vehicles is expanding the approaches that can be exploited to tackle traffic congestion, particularly in urban regions. Connected vehicles pave the way to centralised real-time re-routing, where a urban traffic controller can suggest alternative routes to be followed in order to reduce delays and mitigate congestion issues in the network. In this work, we introduce a centralised architecture and we compare in simulation a number of approaches that can be exploited for re-routing vehicles.

Effective Real-Time Urban Traffic Routing: An Automated Planning Approach

“…It should come as no surprise that the default leads to the worst possible performance with regards to all the considered metrics. Table I also shows the performance achieved by the best Centralised and Decentralised approaches, as for [14]. The presented results suggest that the proposed planning-based approach can effectively route traffic in the modelled region.…”

“…A recent comparative study has considered various optimisation metrics, both centralised and decentralised [14]. The results of the study indicate that the most promising optimisation metric is road occupancy, i.e., the number of vehicles that are currently on a given road.…”

“…To provide the planning engine with a likely evolution of the traffic conditions, vehicles that are already in the network are considered as occupying 5 links at a time: the one they currently are, and the next 4. While this value can be set by the user, this specific value proved to work well for centralised routing [14], and has been identified via a set of preliminary tests. Further, it gives a rather pessimistic overview of the traffic conditions, forcing the system to distribute vehicles as much as possible.…”

“…For the sake of contextualising the performance achieved by the proposed planning-based approach, we also report the results obtained by the best performing centralised and decentralised approaches introduced by Vallati and Chrpa and tested using the same urban region [14]. We will use the Occupancy (decentralised) and n5-Occupancy (centralised), as defined by the authors.…”