Travel time estimation provides valuable information for traveler routing and transportation scheduling. This paperpresents a methodology to estimate travel time using a regression tree model. Vehicle speed is predicted by theregression tree model and it in turn is used as a proxy to estimate travel time, because historical data on travel time iscurrently unavailable. To maintain stable prediction ability in both free-flow conditions and near-capacity flowconditions on freeways, the regression tree model developed includes thirteen explanatory variables in four types:traffic flow variables, incident related variables, weather data variables and time of day variable. The researchreported in this paper is focused on the I5-I205 loop in Portland, Oregon.
展开▼