Remodeling Generation of Trip Activities in Search of "New" Method of Demand Forecast in Developing Countries: A Case Study of Jakarta Metropolitan Area

抄録

This paper aims at developing better trip production/attraction models by adding new independent variables which were overlooked in SITRAMP, a master plan study for the Jakarta Metropolitan Area, and by taking more convincing procedures in selecting variables from a statistical point of view. OLS regression models were employed for modeling based on the Gauss-Markov conditions.
In most of the new models which are estimated by purpose and income group, the number of independent variables has increased after careful selection of variables based on their significance. Also, automobile ownership is added to the models to be used as an important input to the production/attraction models.
Although further research is necessary for actual application of the models, it is hoped that the revised models will contribute to improvement of the first step of the travel demand modeling, which is a "key" process that affects the methods and outputs in the subsequent steps.