Computational experiments for studying impacts of land use on traffic systems

This paper investigates the impacts of land use on traffic systems through computational experiments based on TransWorld, a computational platform for artificial transportation systems. First, we introduce the components of TransWorld and their functions. Second, for given a road network, we employ factorial design and regression to analyze whether and to what extent that the factors of land use have impacts on traffic systems. Three factors, i.e., residential density, distribution type, and attraction ratio, are involved in the related computational experiments during a period from 6:00am to 12:00am. The average speed of vehicles on the road network is used as the performance index to evaluate traffic conditions. Computational experimental results have indicated that all three factors and their interactions have significant effects on traffic systems.