Title

Transferability Of An Intermodal Freight Transportation Forecasting Model To Major Florida Seaports

Abstract

Seaports are important economic generators, and identifying necessary infrastructure improvements is essential to accommodate potential growth at these intermodal facilities. The ability of heavy trucks to access a port's freight terminals is one such operational improvement that needs to be addressed. Freight activity from Florida's major seaports generates more than 10,000 trucks per day. Efficient accessibility to freight terminals and storage facilities at ports can be provided by identifying needed improvements in transportation operations and by developing truck trip generation models to forecast truck trips in and out of the ports. In July 2001 an artificial neural network (ANN) model was developed successfully by using backpropagation techniques to simulate the transportation of freight by heavy trucks generated from an intermodal activity center such as a seaport. The general methodology for developing the model was applied to the Port of Tampa and Port Canaveral in Florida to test the transferability of the ANN modeling technique. From daily vessel freight data, models for both ports were developed successfully and validated at the 95% confidence level with data collected from the field. The validated models were executed for short-term forecasts of truck trips at both ports. Port of Tampa was forecast to have a 0.31% average annual decrease in heavy trucks, attributed to the decreasing trend in bulk commodity shipments. A 5.07% average annual increase was forecast for Port Canaveral, which correlated with historical trends and future estimates for freight activity. Output from these models can be directly used as truck variable inputs to route assignment models used by local and regional government agencies.

Publication Date

1-1-2003

Publication Title

Transportation Research Record

Issue

1820

Number of Pages

36-45

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.3141/1820-05

Socpus ID

0346964050 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/0346964050

This document is currently not available here.

Share

COinS