| As the evaluation index of the urban traffic efficiency, the traffic congestion index (termed as TCI) has been widely used in the management of the urban traffic. Although the effect of the congestion intensity on emissions has long been studied by researchers, few efforts have been made to analyze quantitatively the relationship between the traffic congestion and emissions in an urban traffic network. Some studies have indeed targeted the urban traffic networks, but failed to study the quantitative relationships between the TCI and the vehicle fuel consumption and emission intensities. In this context, the research in this thesis is intended to develop the relationships between the Traffic Performance Index (TPI), a Chinese version of TCI specified in the national standard GB29107-2012 titled Description of Road Traffic Information Service and Traffic Condition, and fuel consumption and vehicle emissions.First, the thesis synthesizes the state-of-the-art regarding the traffic congestion index, vehicle emission model, and the relationships between the congestion and emissions. It further conducts the literature review on the emission uncertainty analysis and the clustering analysis methold, upon which, the research scope, objectives and the approach are determined.Second, the Floating Car Data (FCD) collected in Beijing are processed into two types of data format, the TPI and speed distributions for different classes of roads, and the TPI and speeds for different districts. It is found that there exists an uncertainty of the relationships between TPI and the expressways, arterials, and collectors. As a result, the deviation ratio index is presented to measure the uncertainty.Third, since the uncertainy of the relationship between the TPI and the speed distribution will lead to the uncertainy of the estimation of fuel consumption and vehicle emissions, the study analyzes factors that influence the speed distributions in the traffic network. By means of the K-means clustering method and the Silhouette-based optimal clustering function, it designs a clustering approach for the speed distributions of TPI in urban traffic networks. The research develops two criteria, the Coefficient Variation of the Speed Distribution (CVSD) and the Coefficient Variation of the Emission Factor (CVEF), to measure the quality of the clustering. It is found that the clustering of the TPI-specific speed distributions by days (workdays vs. weekend), time periods (6:00-12:00,12:00-22:00,22:00-6:00), and road classes can substantially reduce the uncertainty of the TPI-specific speed distributions and emission factors.Fourth, by analyzing the calculation mechanism of the TPI and the emission factor, it selects the speed as the connection variable in establishing the relationship model between the TPI and the fuel consumption and the emission factor. Based on the TPI and the speed distribution data after the clustering in a combination of the massive traffic activity data collected by GPS and emission data collected by the Portable Emission Measurement System (PEMS), the relationship curves and the fitting fuctions between the fuel consumption and emission factors and TPI for the entire traffic network, expressway, arterial as well as collector roads are established, and the respective uncertainties are estimated.Finally, the proposed models between the fuel consumption and emission factors and TPI are applied to a case study in Beijing. The trends of the speed as a function of TPI are developed for Chaoyang district, Haidian district, Fengtai district, Shijingshan district, Dongcheng district and Xicheng district. The relationship between the fuel consumption and emission factors and the TPI for the entire network is established. The uncertainty of the fuel consumption and emission factors and TPI for different districts is estimated.The effort in this research can help traffic management authorities acquire a practical approach to evaluate and understand the effects of traffic congestion mitigations on emission reductions. |
PDF Full Text Request