conflated onto the network. Unlike lane count and AADT, no adjustments were needed to the speed limits, as the speed limit values were not affected by differences in data representation. Any segments that were still unknown had their values filled based on the most commonly occurring values for segments of the same name and functional classification. Like lane counts, this process excluded residential streets due to potential outlier observation effects. Any residential road without a known value was assigned the regulatory default of 30 mph. Intersection Control Control was assigned to intersections from several data sources. First, the City of Fort Worth maintains a dataset of signalized intersections. This dataset represented overall intersections that were signalized, so they were simply assigned to the closest intersection location within 30 ft. Next, was the city’s traffic signal dataset. However, this dataset contained points for each traffic signal at an intersection, meaning that a single intersection had multiple points. To simplify this, traffic signals were grouped into clusters based on the name of the intersection they participated in, and the centroid of the cluster was obtained. With this, each cluster was represented by a single point, and was assigned to the closest intersection within 30 ft. Finally, OSM data was used. This data contained both signalized and stop (all way only) controlled information. Again, the control information from this dataset was assigned to the closest intersection within 30 ft.
56
Made with FlippingBook - Share PDF online