Travel time reliability considerably influences travel routes chosen, and the latest update of the travel forecasting model for the Wheeling metro area incorporates measures of such reliability developed from “floating-car” travel time surveys (which also includes a dynamic assignment approach embedded within the standard four-step model process). Presented on a “compare and contrast” basis will be two different travel path estimation procedures, different equation coefficients to represent the variability of travel time at the street segment level (derived from travel time studies both within the Ohio river valley and from England), and different values for the “reliability ratio” (the amount of weight given to the variability in travel time when estimating travel routes taken).
Use of these procedures for travel path selection requires good estimates of modeled travel time under traffic. The overall traffic modeling philosophy utilized follows that previously developed and reported by the author for other small metro areas, and comparable levels of success with previous work will be presented. It will also be shown that model setup effort and computer runtime is only marginally higher than if reliability considerations are ignored.
Future direction will focus on two areas – use of the procedures with commercial GPS travel time data (and differences between sources of data to be mindful of), and extension of the procedures to include railroad grade crossings (in some regions the largest source of travel time variability).