This research group has the experience of applying advanced mathematical models such as discrete choice, latent class, artificial intelligence, structural equation model, and time series models to different data sets mostly transportation data that usually focused on traveler behavior analysis and forecasting future travel demand.
A copula-based joint model was employed to explore the interdependency between destination and departure time choices. The destination choice modeling is developed using multinomial logit model and a binary logit model is used for modeling departure time choice. To obtain better-fitted model several copula functions are used thereafter the frank copula is selected for the final model. Results show that there are some common unobserved factors between these decisions by estimating copula dependence parameters with high statistical significance. Furthermore, there are some commonly observed factors, such as socio-demographic and travel characteristics that appear in the utility functions of both models.
In another study, copula-based joint model is utilized to identify and analyze common simultaneous influential factors on joint car ownership, number of autos and fuel type decisions in Munich
For example, a joint model to predict Munich family car ownership and the fuel type is recently developed by our collaboration with the modeling spatial mobility research team of Prof. Dr. Rolf Moeckel at Technical University of Munich. The paper on this model will be presented at mobile.tum 2018 in Munich and may be in IATBR 2018 in Santa Barbara, California, US. We also developed some copula-based models for different decision travelers made on their trips such as simultaneous decision for departure time and mode, departure time and destination and etc. within cities in Iran.
We have good experience in stated preference data collection that will be useful for evaluating different transport demand management (TDM) policies for example we have project with Tehran (capital of Iran) to evaluate the elasticity of demand to some financial policies such as parking fees, fuel price and congestion pricing. The data collected for 3000 travelers using an application designed to be adaptive for creating some stated preference scenarios according to each traveler current trip. The collected data are analysed with different mathematical advanced model including mixed logit, latent class model and etc. using an open source software called Biogeme (http://biogeme.epfl.ch/). We have also similar experience in field of ICT impacts on shopping trips, airline characteristics on passengers’ choice of itinerary, and etc.