Evaluating an extended traffic congestion avoidance algorithm using TAPAS Cologne

Abstract

Traffic congestion seems to be all around ns, making it difficult for us to avoid. Although reducing the amount of vehicles on the road is a very difficult task, bypassing congested areas is sometimes possible. Here we extend Zammit's Traffic Congestion Avoidance System and compare the effectiveness of two distinct re-routing strategies and their strong and weak points in different circumstances. This is done by simulating traffic in the city of Cologne, Germany using data from "TAPAS Cologne" scenario. The main difference between the strategies lies in the number of routes the system is able to choose from. While the one implemented by Zammit is only able to choose from a predetermined set of routes, the second strategy uses the A *Traffic algorithm to suggest a faster way to reach the destination. Therefore while one system relies completely on the user's knowledge of the road map, the other system is itself able to search for routes by considering different roads that allow the user to reach the destination. Various scenarios were run to evaluate the effectiveness of both re-routing strategies. These consisted of various vehicles leaving from the same departure point at intervals throughout the simulation and travelling towards the same destination. In the case where vehicles are using the re-routing system that relies on the user's known set of routes, all the monitored vehicles knew about the same set of routes. The results show that when using the system, irrelevant of which re-routing approach is used, in general, vehicles are able to reach their destination much faster than if they had remained on the shortest route. In addition the results show that under these circumstances, the re-routing system where vehicles are re-routed along a pre-defined set of routes perform better than when using the A *Traffic implementation. However further tests should be carried out before more general conclusions can be made.