Methods for Forecasting Freight in Uncertainty: Time Series Analysis of Multiple Factors

摘要

The main goal of this research was to analyze and more accurately model freight movement in Alabama. Ultimately, the goal of this project was to provide an overall approach to the integration of accurate freight models into transportation plans and models in Alabama. The first step in the process was to identify the dependent variable and collect the data necessary to develop the models. Initially, Truck Vehicle Miles Traveled (VMT) was the preferred dependent variable however, data collection revealed that the available VMT was not particularly accurate since it is derived from VMT data for all vehicles and there was no validated method for estimating the percentage of trucks in any one year. Therefore, the research team determined that annual Diesel Tax collections would be a good surrogate for Truck VMT. The Diesel Tax collections were used to estimate the Diesel Gallons Sold each year by dividing by the tax rate for that year. This variable, Diesel Gallons Sold (DGS), has the advantage that it could be used to estimate annual truck volumes based on estimate mileage performance for trucks. Thus, DGS was chosen as the dependent variable for this study.