Heterogeneous Traffic Flow Model of Freeway Weaving Sections for Ramp Metering Control

Abstract

Faced with traffic jams on freeways, one of the hottest issues is Dynamic Traffic Management, which manages the freeways dynamically according to real time traffic conditions. However, the controllers use traffic flow models that usually ignore or simplify the heterogeneity effects. Particularly in ramp areas, a traffic flow model considering vehicle types is much more important. The frequent lane changings near on-ramps and off-ramps result in bottlenecks on freeways. Besides, it seriously decreases the capacity of the ramp-weaving sections that the trucks prefer to travel on the outer lanes. To set a proper capacity, which is deduced from the traffic flow model, is the key factor in deciding the ramp metering rate. Therefore, the focus of this research is to develop the heterogeneous traffic flow model of the ramp-weaving sections. The main focus of this research is to clarify the heterogeneous vehicles influence to the areas around them including the lanes they belong to as well as next to them. The characteristic of the truck influence on traffic flow is reflected as the increase on the mean time gap T. Then apply the heterogeneous traffic flow model to ramp metering in order to verify the necessity of the heterogeneous traffic flow model and point out the risk of using traffic flow models that ignore vehicle types. The result came out to be that slight congestion happens in the ramp-weaving section if the optimized ramp-metering rate with the homogeneous model is applied. But more critically, serious congestion happens in the areas before on-ramps. NGSIM U.S. 101 Data is used to verify the characteristics and embody the modified model. This research proves that it is necessary to use traffic flow models that consider the heterogeneity of the traffic flow. The benefits of the heterogeneous traffic flow model will come to light with its development.