Analysis And Research On Urban Road Traffic Emission Based On Sparse Data Reconstruction

Road traffic emission has become one of the main sources of urban air pollution,and highprecision vehicle status data is the core element of micro-emissions analysis.However,in practice,due to the limitations of communication rate,data storage and equipment cost,it is difficult to guarantee the high-precision state data of large-scale vehicles.Therefore,this paper aims to the problem of micro/ mesoscopic road traffic emission based on sparse trajectory data.The paper extracts NGSIM high-precision data set by sparse sampling method,considers the correlation principle between the random characteristics of vehicle acceleration and vehicle state,proposes the velocity reconstruction method based on random acceleration and Markov,and uses NGSIM high-precision data and other two reconstruction algorithms as the evaluation benchmark of the performance of velocity reconstruction algorithm in this paper,and reconstructs a main road of Xi’an city with real Xi’an floating vehicle sparse data.The vehicle velocity of main roads vehicles are used to quantify the microscopic traffic emissions of Xi’an main roads and the mesoscopic traffic emissions of Xi’an city using MOVES model and VSP model.The specific research works are as follows.(1)A wavelet transform-based vehicle velocity/acceleration filtering method is proposed for problems such as measurement errors and outliers in the NGSIM data set.The paper adopts Lankershim Street data in NGSIM database as the data source,analyzes the abrupt change characteristics of vehicle speed and acceleration,initially eliminates the first type of outliers and the second type of outliers in the speed and acceleration curves,secondly decomposes the speed and acceleration to get the scale coefficients,and soft thresholds the inverse transform to get the noise reduction data,numerical analysis experiments with root mean square error and error The numerical analysis experiments are conducted to verify the accuracy of the filtering method.(2)For the velocity reconstruction problem of sparse data of vehicles,a velocity reconstruction method based on random acceleration and Markov is proposed.The data set is sampled sparsely,and the acceleration distribution models of different speed intervals are established according to the random acceleration characteristics of data.The correlation principle between vehicle states is considered,and the missing data is calculated by combining the interval velocity points and Markov transfer matrix,and the velocity reconstruction method based on random acceleration and Markov is established.The paper conducts velocity reconstruction experiments on classical cubic spline interpolation velocity reconstruction and Kalman velocity reconstruction methods with the same data,and conducts experimental comparisons by root mean square error,Pearson correlation coefficient,signal-to-noise ratio,box line diagram and other indexes,and the results show that the algorithm of this paper has better reconstruction accuracy compared with the other two algorithms.On this basis,the sparse speed segments V1 and V2 of the roads in a main road of Xi’an are selected to carry out the real vehicle speed reconstruction experiments and establish the high-precision vehicle speed data for emission calculation.(3)A method of analysis and evaluation of urban traffic emission based on MOVES model and VSP model is proposed for the micro/ mesoscopic emission of road traffic based on sparse data.The paper analyzes the road traffic conditions in Xi’an based on Xi’an floating vehicle data,calculates the VSP distribution of large-scale floating vehicles,localizes the MOVES model and determines the MOVES model parameters,and quantifies the microscopic traffic emissions of main roads in Xi’an based on the speed reconstruction data.Meanwhile,we quantified the mesoscopic road traffic emissions in Xi’an based on the floating vehicle data,analyzed vehicle speed variation trends by means of spatial and temporal distribution maps of floating vehicle trajectories and speed distribution maps,calculated and analyzed the vehicle pollutant emissions per unit time in Xi’an,the emission percentage of total emissions per unit time,and calculated the road traffic pollutant emissions and driving distance the characteristics of road traffic pollutant emissions per unit distance were calculated.The results of the analysis of road traffic emission of Xi’an city show that the floating vehicle speed in Xi’an city is mainly concentrated in the range of 20～45km/h,among which30～40km/h accounts for the highest percentage.The high vehicle emissions are mainly concentrated in 7:00～20:00,with the highest pollutant emission percentages at 8:00 and 9:00a.m.The emission percentages of CO,CO2,HC and NOx are 8.6% and 7.8%,11.1% and 8.8%,7.14% and 7.14%,12.3% and 10.7%,respectively.