실제 도시가로망을 대상으로 한 교통배분 Model의 실증적 검토 = Application of Traffic Assignment Models to the Real Urban Network

다국어 초록 (Multilingual Abstract)

This paper is to study the applicability of the existing traffic assignment models to the real urban network. In this study, ten assignment models and six link performance functions are used to Daegu road network and their estimation accuracy are analyzed comparatively. The results can be summarized as follows. 1) In case of using the same link performance function, there is little difference in the accuracy of estimation among the traffic assignment models. but according to how link performance function is chosen, there is a wide difference in the estimation accuracy among these models. 2) Iterative Assignment model by BPR shows the best estimation result. The next order is followed by Frank Wolf model, Wayne model, Incremental Assignment(IA) model with user equilibrium assignment, IA model, Incremental PATS model and the Incremental Burrell model. Therefore, we can assume that BPR or Wayne model with Iterative Assignment is the most appropriate technique for expressing the current traffic behavior of drivers in real network. 3) Among the equilibrium models, IA model is able to be used effectually in case of needing the real time assignment because this model requires a few number of calculations and short assignment time. Judging from the results above, in order to predict the network traffic volumes of future better, studies to improve the accuracy of link performance functions are needed. Furthermore, differently with general knowledge, considering that the accuracy of estimation of the Iterative Assignment model shows more superior result to the user equilibrium model in real network. The current transportation system is needed to be developed into Intelligent Transportation System(ITS)