Applicability Of Dynamic Traffic Performance Index Models Under Different Road Network Attributes

With the rapid development of economy and the continuous urbanization,the problem of traffic congestion has become increasingly common and prominent in urban area.In order to reflect the operational status of urban road traffic,a series of road congestion indexes have been developed.However,the road congestion index is often developed for a particular city or region,while different cities have different network attributes,resulting in differences in operation characteristics of the urban road traffic,which lead to the applicability problem of the road congestion indexes for different cities.In this context,this thesis studies the applicability of widely used dynamic traffic performance index models to cities with different network attributes.First,based on a thorough review of traffic performance index models,existing dynamic traffic performance index models that can reflect the real-time operational status of urban road traffic are divided into two categories,the model based on the mileage fraction of severe congestions and the model based on the travel time ratio.Then,the two types of dynamic traffic performance index models were applied to the same network.Certain differences were observed among the application results,thus the research emphasis of the thesis is identified,namely the applicability of dynamic traffic performance index models under different road network attributes.To this end,a series of preparations have been completed.Based on the city classification and road network attributes,the thesis designed two parameters for typical cities,the cities’ road network length and road grade configuration.Then,by taking the road network of Beijing as the overall sample,the thesis constructed virtual road networks for different cities using the random number generation method and the determination method for the minimum sample size to provide the network basis for the study.Furthermore,based on the analysis of Floating Car Data(FCD),three kinds of traffic operational status were selected,which provides the traffic operational data basis for the study.In order to improve the application efficiency of dynamic traffic performance index models,the thesis used SQL server to program the models,which provides a means of analog computation for the study.Based on the above preparations,two kinds of the dynamic traffic performance index models were applied to the selected FCD and the constructed virtual network.Based on the signal processing methods and indexes of signals,two single indexes,the sensitivity index(SI)and the stability index(FI)were proposed for evaluating the applicability of the dynamic traffic performance index models.By a weighted average of the standardized single indexes,the comprehensive evaluation index(CEI)was derived.Subsequently,the discriminant method to determine the applicability of the dynamic traffic performance index models was proposed.Finally,the feasibility and effectiveness of the proposed applicability analysis method for dynamic traffic performance index models were validated by case studies.