Patricia : predicting activity-travel interdependencies with a suite of choice-based, interlinked analyses

The development and performance of Patricia, a suite of (choice) models that can be used to analyze and predict activity-travel patterns, is reported. This suite of models, which differs from similar sequential utility-maximizing models of activity-travel patterns in that it incorporates a larger number of choice facets and choice options, was sequentially estimated with activity-travel data collected in the region of South Rotterdam, the Netherlands. The results of the estimation are satisfactory at the level of aggregated origin-destination matrices and individual activity patterns.