Multi-level Models To Measure Road Network Operational Conditions For Megapolises

With the rapid growth of urbanization and motorization, traffic congestion has become a severe problem faced by the development of modern cities. Due to the complicated network structures and the large travel demands, traffic congestion in megapolises is even worse. In order to mitigate the road traffic congestion in urban areas, it becomes extremely important today to measure the traffic operational conditions of urban traffic systems. In this context, the research in this thesis is intended to, on the basis of a comprehensive synthesis of state-of-the-art on traffic congestion measurement methods, develop multi-level models to measure road network operational conditions for megapolises with a focus on the recurrent congestions. The multi-level measurement models, including those at macroscopic, mesoscopic as well as microscopic levels, can be used to measure the operational conditions of the road segment, roads of different classes, as well as the road network.First, in order to measure the operational conditions of road segments, the thesis develops a relationship between the congestion intensity and its influencing factors by using the congestion intensity perceived by the drivers as the dependent variable and the influencing factors as the independent variables, and establishes models for measuring the congestion intensity of road segments based on the cumulative logistic regression method. Then, it presents traffic congestion indices (TCI) for measuring the operational conditions of the road network at macroscopic, mesoscopic as well as microscopic levels. The minimum segment sampling sizes are determined by using the stratified sampling method. Further, by using the comprehensive evaluation method, it employs two variables representing traffic demands from either temporal and spatial perspectives, vehicle-hours-traveled (VHT) index and vehicle-kilometers-traveled (VKT) index, as the weights, and develops traffic demand-based weighting models. After a comparative analysis, it selects the VHT based weighting model as the road network traffic congestion indices model. Finally, in the case study, it applies the proposed models to the road network within the fifth-ring road in Beijing, and completes the evaluation of the road network operational conditions, using the floating car data.