Microscale Spatiotemporal Distribution And Prediction Of Air Pollutants In Urban Traffic Roads

With the accelerating urbanization process in China,the number of motor vehicles has increased sharply,and motor vehicle exhaust has become one of the main sources of urban air pollution.As the direct place for motor vehicle exhaust emissions,urban traffic roads have become the hardest-hit areas of air pollution,which seriously threaten the health of commuters and surrounding residents.Therefore,the study of the spatiotemporal distribution of air pollutants on traffic roads and exploring accurate prediction methods,it has important theoretical guiding significance and application value for assessing road air pollution level and formulating management and control strategies.This paper takes the main air pollutants of traffic roads in Hangzhou as the research object.With the help of air pollution sensor on mobile platform such as electric vehicle and unmanned aerial vehicle(UAV),the horizontal and vertical mobile monitoring of urban traffic road was carried out,and the data of high space-time resolution was collected,the microscale spatiotemporal distribution law of air pollutants in traffic road was explored.At the same time,combined with the data of fixed monitoring station,a prediction model of air pollutant concentration based on classification and regression tree(CART)algorithm was proposed.Finally,the monitoring and analysis system of urban traffic road air pollution was designed and developed.The main research work of this paper is as follows:(1)The overall design scheme was determined by through the demand analysis of the air pollution sensor.According to the overall design scheme,the main function modules are selected,and the parameters of each module are configured.Then,the control flow of the sensor was designed.Finally,complete the development of the sensor;(2)The sensor was carried on the electric vehicle for horizontal mobile monitoring,and the monitoring data with high space-time resolution are obtained.Combined with the road environment,the spatiotemporal distribution of air pollutant concentration at different times of the day and the spatiotemporal distribution of air pollutants under complex road environment were analyzed.And using geographic information system(GIS)technology to realize the visualization of the concentration distribution of road air pollutants.The vertical monitoring experiment was carried out by using UAV,the pollutant concentration data in the range of 100 meters above the road were collected,and the vertical distribution of air pollutants on both sides of in the center of the road was analyzed.(3)Take nitrogen oxides(NO_X)from air pollutants as an example.The influence of traffic flow,wind direction,wind speed and other factors on NOx concentration was analyzed,and a concentration prediction model based on CART regression tree was constructed.Taking a certain area of Hangzhou as an example,the model constructed by the application was verified.The result shows that the prediction coefficient of the prediction model was above 0.92,which was better than Back Propagation(BP)neural network and support vector machine.The applicability of the model under different conditions was analyzed,and it was verified that the prediction model can better adapt to the forecasting requirements under different conditions.(4)The overall design scheme was determined by through the demand analysis of urban traffic road air pollution monitoring and analysis system.Based on data such as sensors,fixed monitoring stations and electronic bayonet,combined with algorithms such as global positioning system(GPS)data missing completion,air pollutant concentration prediction and other algorithms,the urban traffic road air pollution monitoring and analysis system based on client/server(C/S)architecture was developed.And the functional modules and implementation process of the system were elaborated.