Evaluation of Intersection Properties Using MARS Method for Improving Urban Traffic Performance: Case Study of Tekirdağ, Turkey

Abstract

Increasing urban traffic performance is a technical problem that has been investigated by many researchers these days. Traffic performance can be increased in many ways, as part of the transportation planning process, on smaller scales, or with different methods and techniques. The determination of traffic intervention areas in urban transportation planning is an intervention type that determines the rate at which the traffic performance will increase. Although transportation planning is an integrated issue, the type of traffic modification and prior intervention on intersections are often determined with partitive paradigms and strategies. It is a significant opportunity for decision makers to be informed in advance of the effects of intersection characteristics on the overall traffic performance. However, it is not an attempted or tested concept to perform a general assessment of the impact of the intersection characteristics on the overall performance of the intersections. In this study, a four-stage integrated analysis including the multivariate adaptive regression splines (MARS) method is proposed for the overall traffic performance evaluation. The traffic characteristics of intersections are first indexed and categorized. The intersection performance results are then obtained using the VISSIM traffic simulation software. Subsequently, the relationship between them is determined with the MARS method, and the effects of the intersection characteristics on the traffic performance are investigated. Tekirdağ, a metropolitan city located in northwestern Turkey, is selected for the case study. According to the obtained simulation results, the suggestions related to the development of intersection performances are made and tested.