Smart algorithms for modeling urban traffic

Densely populated areas, such as Singapore, face the problem of traffic congestion on daily basis. Of late, route guidance has become a feasible and emerging solution to diminish the congestion problem. The route guidance systems are utilized to find the optimal solution by finding the shortest paths (or minimum travel time) from the origin to destination. There are two types of route guidance systems: static and dynamic, hi the static route guidance system, the shortest path does not depend on the changes in the ground traffic conditions. However, in a dynamic route guidance system the shortest path is updated regularly based on the future traffic conditions. Moreover, due to traffic congestion, it is not feasible to provide the same path to all the vehicles. Hence, we use k-shortest (multiple) paths to tackle this problem. Additionally, the Land Transport Authority (LTA) of Singapore provides speed data required for routing in the form of speed bands, which has limited (low) resolution. Therefore, the purpose of this dissertation is to compare the efficiency of this low resolution data with the high resolution data. We do this by estimating the travel times and the shortest paths for both the high and low resolution data, and compare them. From the results obtained we conclude that, since the differences in the travel times, and the changes in the shortest paths are negligible, the speed band data provided by LTA can be used in the route guidance systems to determine the optimum routes and the shortest travel times as a solution to the rapidly increasing congestion issues in Singapore.Master of Science (Computer Control and Automation