The use of smart card data in public transport assignment models

The smart cards that are used in the transit systems in many metropolises, provide an opportunity to predict the travel characteristics like time intervals, Origin-Destination (O-D) pair loads, mode choice, transfers etc. by following the users at the individual level, especially for regular (the trips by the same user, in the certain days of the week, performed between the same pair of origin-destination) travels. Main question whose answer is sought in the study is whether a result which is as reliable as that of the household travel survey (HTS) can be obtained when the public transport O-D matrix is estimated with the smart card data. In this context, in order to analyze of Izmir Smart Card Fare Collection System Data within the scope of thesis study, "Origin-Destination Matrix and Passenger Route Choice Model with Smart Card Data" and "Reliability Criteria Model for Bus Lines" were developed. In addition, the similarities and differences between public transport O-D matrix that one of the important output aimed at public transportation demand models, obtained from smart card data and transportation household survey results was analyzed. In accordance with this aim, the passenger route choice behaviors obtained from the smart card data have been examined over a long period of time and assignment results obtained from the Spiess and Florian (1989) strategy-based assignment model were compared with smart card data. Finally, linear regression, exponential and logistic regression based route choice models were developed and a method proposal has been developed aiming increasing the efficiency of the results obtained from public transportation assignment models with uof passenger route choice behaviors to be obtained from smart card data