Direct Demand Forecasting Model for Small urban Communities Using Multiple Linear Regression

Abstract

Forecasting traffic volumes to support infrastructure decisions is the heart of the travel demand modeling process. The most commonly used methodology for obtaining these forecasted traffic volumes is the four-step process that considers generation, distribution, mode choice, and route assignment of trips. Each step of the process is performed independently, almost always through the use of computer software, to achieve the final traffic volumes. This paper examines the possibility of forecasting traffic volumes by using a multiple linear regression model to perform what is termed direct demand forecasting. The direct demand forecasting model generates traffic volumes for roadways through the development of a functional relationship between roadway characteristics and socioeconomic influences. A direct demand travel forecasting model has been developed and applied, with a small urban area as a case study community. Results are consistent with those obtained from the traditional four-step methodology.