نوع مقاله : علمی - پژوهشی
نویسندگان
1 استادیار، دانشکده پژوهش های عالی هنر و کارآفرینی، دانشگاه هنر اصفهان، اصفهان، ایران
2 استادیار، گروه اقتصاد، دانشکده علوم اداری و اقتصاد، دانشگاه اصفهان، اصفهان، ایران
3 دانش آموخته کارشناسی ارشد، گروه اقتصاد و کارآفرینی ، دانشگاه هنر اصفهان، اصفهان، ایران
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Since there are different types of transportation modes and their corresponding time cost in cities, one way for minimizing transportation cost is finding the best transportation mod out. In this paper, the factors influencing the choice of trip mode in the framework of the polynomial logit model were investigated. 384 data were obtained by random sampling of intra-city trips to Azadi Square in winter of 2017 in Isfahan. The dependent variable (traveling mode selection) has four floors: personal car, regular bus, express bus and taxi.
The personal car was selected as the reference class and the other classes were compared with the reference class. According to the results, in the Logit Model, between the reference alternative and the conventional bus, the cost variables, walking time and expectation of a negative relationship and lack of access to the fastest bus, personal ownership of the car and travel time have a positive relationship with the possibility of choosing a typical bus compared to the reference alternative. In the second Logit model, between the alternate reference and the taxi alternative, the travel time, walking time and expectations of the negative relationship and the cost and ownership variables are positively correlated with the probability of choosing a taxi mode than the reference alternative. In the Logit III model, there is a positive relationship between the alternate reference and the bus of the fast-moving travel time variables, the lack of ownership of the vehicle and access to the fast-moving bus, and the cost variables, walking time and waiting time are negatively related to the possibility of choosing a bus to the alternative than the reference. The results of clustering showed that clustering increased the prediction accuracy of the model relative to the ungrouped state.
کلیدواژه‌ها [English]
)
