Abstract
Travel time functions, or volume delay curves, define a relationship between the volume of traffic on a road and the travel time along the road. While procedures for calibrating travel time functions from travel time and volume data have been developed, and options for estimating these functions for urban roads have been outlined, there are virtually no guidelines for the data collection process.
The study reported here provides insight into the consequences of errors in the parameters of travel time functions. The results of this study have application in the development of data collection guidelines for travel time functions.
The paper begins by discussing the econometric estimation of travel time functions, identifying the various types of estimation error and outlining the reasons for selecting the effects of parameter bias as the focus for this study. A simple network is used to visualise the effects of parameter bias before an empirical study is discussed which uses a network representative of those encountered in practice. Finally, the conclusions of the study represented and directions for future research are identified.