The multiple discrete-continuous extreme value (MDCEV) model is an advanced model used to estimate activity duration. It contains three parameters: baseline utility, translating satiation, and pure satiation. The translating satiation parameter is expected to capture the constant marginal utility effect, but it does not. Therefore, a modified model was developed that adds a fourth parameter (the power parameter) to the translating satiation parameter to capture the constant marginal utility effect. In addition, this research applies the power parameter to either and both the translating satiation parameter and the pure satiation parameter to examine effects of the parameters’ interaction to further improve the MDCEV model’s accuracy. The proposed model was applied to data sets from two countries, Saudi Arabia and Germany, to test the applicability of the modified model to any data set. This research found that adding the power parameter exclusively to the translating satiation parameter was the best model structure to maximize the accuracy of the MDCEV model for both data sets. Because an activity duration model is part of an activity-based travel demand framework, this improvement will lead to better predictability of activity-based travel demand. As a result, transportation planners can make appropriate decisions regarding future transportation infrastructure projects, which in turn will lead to a reduction in costs associated with these projects and decreased delays for transportation system users.
The main step in estimating travel demand is establishing the modeling framework. Conceptual travel demand modeling frameworks for mega-events were established based on a literature review of frameworks and the studies that form the skeleton of these frameworks. Studies performed on mega-events demonstrated the importance of modeling mega-events separately from regular daily activities. Studies conducted on market segmentation shed light on the importance of modeling mega-events participants separately from nonparticipants and have resulted in improved guidelines for mega-event host cities aiming to reduce road network congestion. Criteria were established to select statistical software suitable for project inputs (e.g., project size). Finally, the estimated modified MDCEV model was generalized for use in transportation planning around the globe where limited planning models are available.