Wang, M., Pougala, J., and Bierlaire, M. (2023)

Initial comparisons between Multiple Discrete-Continuous Extreme Value (MDCEV) model and Optimization-based Activity Scheduling Integrating Simultaneous choice dimensions (OASIS) framework

23rd Swiss Transport Research Conference (STRC), Ascona, Switzerland

In activity-based travel demand modeling, daily scheduling process is a critical module which can help researchers understand how travel demand is generated from a micro perspective. Multiple Discrete-Continuous Extreme Value (MDCEV) model and Optimization-based Activity Scheduling Integrating Simultaneous choice dimensions (OASIS) framework are both widely used activity scheduling models in recent years. This paper takes an initial attempt to compare the two models by evaluating the forecasting performance in activity participation and activity duration choice dimensions. We estimate the models by the same samples of Lausanne from the 2015 Swiss Mobility and Transport Microcensus (MTMC) and carry out model simulation to test the ability of reproducing real data in both aggregate level and disaggregate level for the models. Because of the limitation to integrate time of day decisions of MDCEV model, more detailed comparative study of the two frameworks will be conducted when the MDCEV model is combined with scheduling model in the future.