This study developed a multi-facted procedure to forecast traffic volumes on urban segments of the state and interstate highway systems for the Indiana Department of Transportation. A detailed statistical analysis was conducted to develop alternative models to forecast traffic. The models and the procedures described in this study's Final Report are intended to provide highway planners with a tool for simple, fast and inexpensive estimation of traffic projections. Emphasis was also placed on achieving reliable input information for the forecasts of any background factors (independent variables) used in the models. This study improves on previous efforts to build simple and reliable procedures to forecast traffic. Among the several models examined in this study, elasticity and 1ag-AADT models were finally selected because of their simplicity and better performance. This study has employed several measures to evaluate the forecasting models. Often, combined forecasts developed in this study were found to perform better than the forecasts from any of the constituent models. Some problems and limitations of the models -- and suggestions to overcome them -- are discussed.

Report Number



traffic volumes, forcasting, models, statistical analysis

Performing Organization

Joint Highway Research Project

Date of this Version