Modeling And Simulation Of Destination Choice Behaviors Based On Space-time Prism

Activity-based model is an efficient technology to analyze residents'trip characteristics, and evaluate the application of traffic demand policy. Based on the activity analysis which participate acturally, it constructs the model directly using data surveyed which not aggregate by traffic zones. As a complex constituent part of the activiy-travel theory, destination choice is an optimization process for activity schedule under the constraint of space and time. In the exiting traffic planning projects, destination choice behaviors were analyzed centered largely in trip distribution and most used aggregate models, rarely used disaggregate methods. Disaggregate behaviors research of residents'destination choice is of great practical application potentiality, and it has been getting to the attention of scholars, advantages are show below, It can be used to the evaluation of traffic demand management strategies, and applied in the exhaust emission indexes calculation which based on trip distances,also, It could be applied in the interaction studies between traffic demand and land use. Furturemore, with the development of data acquisition technology and GIS, which provide a better basis, the analysis technique of space-time activities and travel behaviors will be promoted and applied widely.Based on these backgrounds and combined with the "11th Five-Year Plan" to support science and technology project, Research on Travel Behaviors Characteristic and Model of Urban Transportation(2006BAJ18B02-03), this paper constrainted the activities scopes in specific prisms by using space-time prism analytical technique, modeled and simulated the destination choice behaviors based on the activity-travel schedules. The results enrich the content of disaggregate modeling technology, and would be helpful to promote the application of activity-travel theory in the destination choice behavior analysis and application.The main contents of this paper can be summarized as follows:(1) Activity-travel behavior analysis based on space-time prismIn this section, activity-travel theory and its simulation methods were discussed, also including space-time prism analytical technique. On these bases, the analytical framework of activity-travel behavior based on space-time prism was established(2) Activity-travel simulation system construction based on space-time prismConstructed an activity-travel simulation system conforms to the situation in our country based on time continuity. The destination choice model is a basic component of that system; also its validity was tested by simulation experiment. The system structure and module functionality were discussed.(3) Time duration model construction and simulationDistinguish with discrete model of time which usually used. Constructed time duration model and put forward its simulation strategy(4) Modeling destination choice behaviors based on utility maximum theoryModeled destination choice behaviors based on utility maximum theory. A hierarchical modeling strategy was discussed. Considering attracting ratios of zones, constructed the model and identified the parameters of non-liner utility function. Moreover, the results were compared with often-used clogit model.(5) Modeling destination choice behavior based on rule theoryAgainst the deficiencies of the models based on utility maximum theory, from another standpoint, the destination choice model using rule theory was constructed. The joint modeling strategy was discussed, and three types of activity scope category models, Decision tree, Bayes decision network, Bayesian network augmented tree(BNT) were constructed and evaluated. The results show that BNT is the most suitable for the analysis of activity scope.(6) Simulation experiments of destination choice modelsCarried out the simulation experiments of models constructed which based on utility maximum or rule theory. The advantageous and applied limits of these models were evaluated. The results show that the ruled based model is better than utility based by comparing their efficiency and accuracy.This paper aim to model and simulate destination choice behaviors of urban residents based on space-time prism. The main innovations of this paper are:(1) An analytical framework of the activity-travel behaviors based on space-time prism was established.(2) Applied destination choice behaviors in the activity-travel simulation system. Constructed time duration model and put forward its simulation strategy(3) A destination choice model based on utility maximum theory was constructed. Considering attracting ratios of zones, constructed the model and identified the parameters of non-liner utility function.(4) A destination choice model based on rule theory was constructed. The activity scope category models BNT was constructed and evaluated.