Distribution Characteristics And Models Of Vehicle Specific Power Of LDV And HDV For Traffic Emission Estimation

Quantifying fuel consumption and emissions in traffic networks is a necessary way for evaluating traffic control strategies effectively. As the best explanatory parameter of fuel consumption and emissions, Vehicle Specific Power (VSP) has been used widely in the quantitation of vehicle fuel consumption and emissions. At the same time, plenty of data of average travel speed can be obtained from traffic data collection equipment, for example, floating car system, and traffic models. Therefore, it is important to investigate the relationship between these two parameters,(VSP distribution and the average travel speed). However, most recent studies of VSP distribution are on Light-Duty Vehicles (LDVs) for low or moderate speed driving conditions. In this context, this thesis considered Heavy-Duty Vehicles (HDVs) as one of objective vehicle types, and included high speed as an objective operating mode, and then this thesis conducts comparative analysis on the distribution characteristics and models of the VSP distribution of LDV and HDV for traffic emission estimation.First, the thesis analyzes fuel consumption and emission models in which different methodologies are applied to describe vehicle behavior characteristics. The study synthesizes the development of VSP distribution models used on the quantification of vehicle fuel consumption and emissions. The research content of in this thesis is refined based on reviewing the potential shortages of the exsiting studies.Secondly, data collection plan and data processing method are designed for collecting driving behaviors for LDV and HDV. The results are produced with a large number of driving segments in different bins of average travel speeds for LDV and HDV on different road types., The VSP data are also calculated for analyzing the characteristics of VSP distributions in next chapter.Then, the thesis put emphasis on analyzing the characteristics of VSP distributions for LDV and HDV in different speed bins. The study illustrates that, both VSP distributions of the LDV and HDV follow normal distribution in different speed bins, and the value of VSP Bins for the peak of VSP fractions are identical with the VSP value when cruising at the average travel speed. After a comprehensive analysis, regular relationships are found between the VSP distribution and the average travel speed. Mathematical models of VSP distributions for LDV and HDV are then derived for traffic emission estimation. Then the real-world fuel consumptions are used to test the applicability of the proposed VSP distributions model for fuel consumption estimations. It is found that the relative errors of fuel consumption estimation for LDV and HDV are3.4%and3.0%respectively.Finally, aiming at the problem in the existing applications that using VSP distributions of LDV for HDV’s emission estimation, differences of VSP distributions between LDV and HDV are analyzed, and error analysis is conducted for the misusing of VSP distributions. It is found that the peak of VSP distributions of HDV is higher than that of LDV in the same speed bin, especially in high speed range. It is also found that the value of VSP Bins for the peak of LDV’s VSP distributions is always higher than that of HDV, and the difference increases with average travel speed. Then, the real-world and modeled VSP distribution for LDV and HDV, as well as the real-world fuel consumption rate are used to estimate the errors of estimating HDV fuel consumptions. The study result demonstrates that fuel consumption factors of misusing LDV’s VSP distribution are significantly higher than those from the real-world, with an average relative error of24.93%, and the relative error increases with the average travel speed.