Abstract

The focus of this paper is to research the relationship between dependent variables (bus travel delay and travel time per mile) and explanatory variables (road network, bus route and traffic characteristics), and develop models to estimate bus travel time from a bus-stop to the next bus-stop along a route. Three different types of bus travel delay models and one travel time per mile model were developed, evaluated and compared by time period (morning peak, mid-day off-peak, and evening peak) to assess the applicability of these models for improved arrival prediction. The three bus travel delay models are based on free flow travel time, travel time based on the Bureau of Public Roads (BPR) equation (default parameters) and travel time based on BPR equation (local parameters from regional transportation network model). Automatic vehicle location (AVL) data obtained from the Charlotte Area Transit System (CATS) in the city of Charlotte, North Carolina was used to extract data for 81 links along 15 bus routes and develop models using multivariate linear regression analysis. Data for seven links that are not used for modeling were then used for validation. Results obtained indicate that bus travel time per mile model has better predictive capability than the bus travel delay models. The hourly traffic volume, number of bus-stops per mile, signalized intersections per mile and two-way stop controlled approaches per mile tend to have a statistically significant effect on bus travel time and delay.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call