Multipath Assignment Calibration for the Twin Cities

by Jerry L. Edwards, Asst. Prof.; Dept. of Civ. and Mineral Engrg., Univ. of Minnesota, Minneapolis, Minn.,
Ferrol O. Robinson, Transportation Planner; Barton-Aschman and Assocs., Inc., Minneapolis, Minn.,

Serial Information: Transportation Engineering Journal of ASCE, 1977, Vol. 103, Issue 6, Pg. 791-806

Document Type: Journal Paper


Several probabilistic multipath traffic assignments were made to the 1970 Twin Cities highway network. Several aggregate measures were selected to determine the goodness of fit to recorded ground count volumes, including volume-to-ground ratios by route, facility type, screenlines and cutlines; and percent root-mean-square error and percent standard deviation by volume group and facility type. Results indicate that link impedances composed of a linear combination of time and distance and weighted more heavily in favor of distance, when used together with a value of 0.002 for the diversion parameter Θ, produces an assignment that most closely approximates the recorded volumes, and is generally a better assignment than that produced by the all-or-nothing model using calibrated travel times. The probabilistic model exhibits some sensitivity to changes in the diversion parameter Θ, but model sensitivity to changes in the impedance function seems much stronger.

Subject Headings: Travel time | Calibration | Probability | Parameters (statistics) | Traffic models | Traffic assignment | Transportation networks | Aggregates

Services: Buy this book/Buy this article


Return to search