One of the most important issues in urban transport planning is the analysis of travel demand, which can be used to describe the current situation, to predict the travel demand of the coming years, as well as to evaluate transportation policies in the short and long horizons. In order to analyze the demand for urban travel, it is necessary to know the transportation behavior of people. The inaccuracy in estimating travel demand models can lead to the lack of proper policies for controlling traffic flows in the area studied and the risks and costs of irreversible costs for that area.
In this dissertation, disaggregate models for production and distribution trips for the three purposes of work, shopping and education in relation to residents of Isfahan city have been made. These models have been calibrated, using source-destination census results that were taken from the city of Isfahan in 2012. Three types of regression, linear regression, Binary logit and ordered logit are used to make disaggregate trips production models at both individual and family levels. Also, the multinomial logit model has been used in the construction of disaggregate trip distribution models. The results of the making disaggregate trips on the trip production with the purpose of work have shown that employment variables, gender, age, and simultaneous access to personal transport and certification have been effective in generating these trips. The results also show that gender variables, age range, and simultaneous access to personal tra ort and certification for trip production models with the aim of shopping, and student variables and student access, and access to personal vehicles have been effective in the trip production models for the purpose of education. The results of making individual travel distribution models have shown that gender variables, simultaneous access to personal vehicles and certificates, individual household finances, age and type of occupation for travel modeling models with the purpose of individual work and gender variables, simultaneous access to personal transport And the personal and financial status of the household has been effective in shopping distribution models. The results have shown that gender variables, students, or students and simultaneous access to personal tra ort have been influential in trip distribution models with the aim of training. Also, in order to predict the future traffic situation in Isfahan city, aggregate models for the production and distribution of residents' travels in Isfahan, with three goals of work, shop and education have been estimated. In the making of these models, the origin-destination data of 1391 have been used. To make aggregate trips production models, a linear regression model and two variables of population and per capita ownership of a personal vehicle and to make aggregate trip distribution models, multinomial logit model and two variables of origin-destination distance and services and facilities in the destination area has been used.
In the final part of this thesis, the correction of the prediction model of the metropolitan population of Isfahan, which was presented by the Sharif Research Institute in 2000, has been addressed. The model was corrected by considering the factor of access to facilities and services. Comparison of the results of the corrected model and the Sharif model shows the superiority of the corrected model.