Analyze Spatial And Temporal Characteristics Of Exhaust Emissions Around Intersections Based On Taxi Trajectory Data

With the social progress and continues leap of people’s living standard, vehicles have become indispensable means of transport when people travel, and the number of vehicles is growing year by year. Consequently, traffic congestion and emissions pollution increasingly sharpen, and emissions pollution is constantly threatening physical and mental health of residents and comfortable degree of urban environment. Therefore, how to accurately estimate regional vehicle exhaust emission and the spread of exhaust, and how to propose treatment schemes have become the urgent needs of traffic management department and environmental regulatory authority.Fixed or mobile air quality detectors are the main methods of traditional air quality survey. Detectors can accurately calculate the content of NO2, NO, SO2, and other hydrocarbons in the air. However, regardless of the fixed or mobile detections, there are spatial and temporal limitations of detections. In order to solve these problems, we consider starting from the source of emissions, to introduce large amount of existing vehicle trajectory data, derive characteristics of driving behavior from trajectory data, realize inversion of possible vehicle exhaust emission, and then adjust vehicle exhaust emission through data detected by air quality detector, and develop regional vehicle exhaust emission estimation methods. In particular, with respect to the important role that road intersection play in the transportation system, this paper takes intersections as main object of study, and look forward that future research can be extended to the entire road network.The realization of this method consists of two parts. First, vehicle exhaust emissions estimation model is established. We conduct a correlation analysis and regression analysis based on air quality data and volume of traffic flow data collected on scene, and attempt to establish vehicle exhaust emission estimation model. Second, model input parameter is extracted. We develop a method to derive traffic flow information from vehicle trajectory data, including inversion of average delay of traffic flow of different roads around intersections in different periods of time, and inversion of volume of traffic flow around intersections in different periods of time, and we use volume of traffic flow derived from trajectory data as input parameter to estimate vehicle exhaust emission. An application of the method described above is conducted using taxi trajectory data collected in Shanghai in2012. The result shows that the method we proposed can establish a regional vehicle exhaust emission estimation model, and efficiently process large volume of vehicle trajectory data, deriving traffic flow information around intersections to realize the estimation of vehicle exhaust emission. In future studies, we need to consider more factors that affect vehicle exhaust emission, and take them into account of our estimation model to further improve the estimation accuracy.