摘要：AbstractDeveloping province-wide freight demand models has been a difficult task not only for Canadian provinces but also for almost all transportation jurisdictions in the world. The major reason behind this difficulty is the lack of data required to accomplish the task. However, this study has utilized a variety of public sources of data, such as Input-Output Tables and census, together with other borrowed data to develop a freight demand model for the Province of Alberta. Basing on the available data, the commodity- based approach has been adopted. The overall quality indicators for the developed model are 0.87, 8%, 29% and 39% for R2, APE, AAPE and RMSE respectively. When the modelled highways are divided into the Freeway/Expressway and Arterial groups, the resulting quality indicators are as follows: 0.83, 4%, 30% and 42% for R2, APE, AAPE and RMSE respectively for the Freeway/Expressway group and 0.88, 11%, 28% and 36% respectively for the Arterial group. The satisfactory accuracy achieved indicates that it could be useful for a variety of policy analyses involving freight transportation. In addition, using public sources of data minimizes the development cost of such models, which is seen as an important step forward to leverage resources and support the innovations in highway agencies.
