Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Freight-Trip Generation Model

Freight-Trip Generation Model Freight demand models are important analytical tools for estimating demand per retail establishment, which is assumed as the total vehicle arrivals for loading and unloading purposes. Urban freight parking demand prediction is particularly useful in the context of transportation planning to determine, for example, infrastructure options and investment scale (e.g., freight parking). Achieving a balance between the predictive capabilities of the model and the feasibility of application is a challenge, as data regarding the population of retail establishments are frequently minimal. An establishment-based freight survey was used to collect data for 604 retail establishments in the city of Lisbon, Portugal. The relationship of candidate independent variables (e.g., store sales area and supply chain characteristics) versus the total number of weekly deliveries was investigated. Variables were chosen on the basis of suggested choices from the literature but also considering an exploratory perspective. Various sets of variables were modeled under ordinary least squares (OLS) linear regression and generalized linear model frameworks. Several tests were performed to assess model output quality. The analysis of the variables showed that those more commonly used in practical applications were not necessarily the best predictors of demand. The number of employees was consistently a better predictor than the area of the establishment or warehouse. Establishment category was the dominant variable. Supply-chain-related variables slightly improved the predictive capabilities of the models. OLS models showed better predictive capabilities but suffered from heteroskedasticity problems. Overall, the predictive capabilities of any of the models with the chosen methodology were lower than what is considered acceptable for a practical application. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Transportation Research Record SAGE

Loading next page...
 
/lp/sage/freight-trip-generation-model-jLyRFPc6e5

References (20)

Publisher
SAGE
Copyright
© 2014 National Academy of Sciences
ISSN
0361-1981
eISSN
2169-4052
DOI
10.3141/2411-06
Publisher site
See Article on Publisher Site

Abstract

Freight demand models are important analytical tools for estimating demand per retail establishment, which is assumed as the total vehicle arrivals for loading and unloading purposes. Urban freight parking demand prediction is particularly useful in the context of transportation planning to determine, for example, infrastructure options and investment scale (e.g., freight parking). Achieving a balance between the predictive capabilities of the model and the feasibility of application is a challenge, as data regarding the population of retail establishments are frequently minimal. An establishment-based freight survey was used to collect data for 604 retail establishments in the city of Lisbon, Portugal. The relationship of candidate independent variables (e.g., store sales area and supply chain characteristics) versus the total number of weekly deliveries was investigated. Variables were chosen on the basis of suggested choices from the literature but also considering an exploratory perspective. Various sets of variables were modeled under ordinary least squares (OLS) linear regression and generalized linear model frameworks. Several tests were performed to assess model output quality. The analysis of the variables showed that those more commonly used in practical applications were not necessarily the best predictors of demand. The number of employees was consistently a better predictor than the area of the establishment or warehouse. Establishment category was the dominant variable. Supply-chain-related variables slightly improved the predictive capabilities of the models. OLS models showed better predictive capabilities but suffered from heteroskedasticity problems. Overall, the predictive capabilities of any of the models with the chosen methodology were lower than what is considered acceptable for a practical application.

Journal

Transportation Research RecordSAGE

Published: Jan 1, 2014

There are no references for this article.