Citation

Van Aerde M, Yagar S. Transp. Res. A Gen. 1983; 17(4): 301-313.

Copyright

(Copyright © 1983, Elsevier Publishing)

DOI

10.1016/0191-2607(83)90093-6

PMID

unavailable

Abstract

The effects of traffic volume on speeds of 2-lane rural highways are analyzed using a large data bank compiled in Ontario, Canada in 1980. These effects are examined for the entire practical range of volumes using two types of linear models. The first considers volumes of cars, trucks, recreational vehicles and other vehicles in the direction being analyzed, and the total volume in the opposite direction. The second model considers only aggregated passenger car units in the main direction and opposing direction. The 10th, 50th and 90th percentile speeds are estimated using each of the above models, so that the ultimate decision maker can cater to any chosen subpopulation or estimate the fraction of drivers that drive at any given speed. Regression equations are presented for Ontario locations which were analyzed, along with some sample graphs. It is observed that speeds are generally quite insensitive to volumes for a large practical range of volumes, and that the percentile speed curves tend to converge as main directional volumes increase. The 90th percentile speed equation is found to be most sensitive to main directional volume and the 10th percentile equation is least sensitive. The relative sensitivities of the different percentile speeds to opposing volumes are far less pronounced. Speeds are generally less sensitive to opposing volume than to with-flow volume.