Title: Should Larger Time Penalties Be Applied for Modeling Internal-External Trips?
Abstract: This paper reports some findings in the study of use of travel time penalties for distributing internal-external (I-E) and external-internal (E-I) trips. The study was based on a travel demand model that was developed by the authors for Parkersburg-Marietta Metropolitan Statistical Area (MSA), an interstate metropolitan area including a county in Ohio and a county in West Virginia. In the development of the travel demand model, specifically in the estimation of the exponential function parameters of the gravity models for trip distribution, the authors investigated the use of relatively large extra travel times, which averaged 15 minutes, on the trips traveling between internal traffic analysis zones (TAZs) and external stations. The purpose of applying those large penalties was to more accurately reflect the actual travel times of those trips. The penalties were derived from the household travel survey in the same area. As a result, the Root Mean Square Error (RMSE) of system-wide traffic assignments revealed that the application of large penalties improved the model, compared with the situations where no penalties or small penalties (2-5 minutes) were used. Inspired by this interesting finding, the authors further investigated the trip matrices generated by the gravity models and found noticeable differences between the scenario with large penalties and the one with no penalties. Based on the findings, the authors discussed in a broader range the I-E and E-I trip distribution issue for mid- or small-sized communities where decent proportions of I-E and E-I trips exist, which may be otherwise distorted in a travel demand model without careful handling. The authors finally suggested larger travel time penalties should be applied and an approach to estimating appropriate travel time penalties was also proposed when no survey data is available.
Publication Year: 2004
Publication Date: 2004-01-01
Language: en
Type: article
Access and Citation
Cited By Count: 1
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot