Last Updated: 7-2017
This study mainly develops and applies the shortest path approach based on time and distance for route choice analysis of car drivers. Person-based GPS data is analyzed in an attempt to develop a more practically accessible approach that can be used, for the time being, in the place of the GPS tracking approach to estimate, to the best possible extent, the parker’s route choice in a timely and costly manner. GPS data was extracted from the B-Riders stimulation program which took place in the province of Noord-Brabant, the Netherlands. The study presents the applied methodology, the way the data has been collected, organized and analyzed, and the findings of the study. The main hypothesis being tested in this study is that car drivers choose routes in order to minimize either the total travel time or distance for “non-daily shopping” trip purpose. All the trips have a destination to the central area of Tilburg, the Netherlands. A GPS data set of 83 car trips is analyzed in TransCAD (GIS). Combining GPS and GIS can provide great insights into car drivers’ route choice decision making. The comparison of the observed routes to shortest alternative paths based on time and distance revealed that observed routes are significantly longer than the shortest paths. However, it appears that shortest path routes based on time correctly identify on average 75% of the observed routes’ total travel distance compared to 52% in the case of shortest distance routes. These findings suggest that drivers consider time more than distance as their path selection criterion. It is also implied that more research is required to define if other factors affect parkers’ route choice. In addition, there is also space to improve as well as to reconsider the way travel time and distance attributes were expressed in this study. Finally, the practical application of the shortest path approach based on time is demonstrated through a traffic assignment model for a specific parking facility. The estimated traffic flow was compared to the actual traffic flow. The performance of the shortest path algorithm considering travel time shows good potential for further development. Therefore, more data is required in order to provide robust evidence about the performance of the traffic assignment model and the magnitude of the possible shortcomings of its application.