Traffic Noise Prediction Model And Planning Control Strategy Based On Urban Multi-source Open Data

Traffic noise is the main source of urban environmental noise and has become a serious pollution factor affecting the quality of life.Urban planning is a comprehensive deployment of regional functions,which is closely related to traffic noise pollution.However,the existing noise research is mostly limited to the micro level,which leads to the lack of effective prediction of traffic noise and the lack of prevention and control technologies in the field of urban planning.In view of the limitations of traffic noise prediction and control,European and American countries have adopted overall planning from the macro level,while China’s research on urban traffic noise has just started,but it has also begun to realize that"reasonable regional planning is the fundamental way to solve the problem of traffic noise pollution".Taking Dalian city as an example,this paper introduces the concept of urban planning into the study of traffic noise prediction and control.Based on the grid scale of 300m×300m under the block level,the building data,road data,interest point data and vitality value data are screened out from numerous open data as the research basis.The building data and road data are used to draw the research model,and the accessibility analysis of Depth Map space syntax is carried out on the road data.Based on the GIS spatial clustering analysis of the points of interest data and the vitality data,the influence of multi-source open data on traffic noise was explored through the combination of the two.Then,the traditional econometric model,the geographically weighted regression model and the spatial econometric model were established respectively to carry out the prediction of traffic noise and the research on the influencing factors.The conclusions are as follows:(1)There is spatial aggregation of urban traffic noise at the block level.As the spatial location approaches,the aggregation becomes more obvious and the correlation becomes higher.According to spatial autocorrelation,there are four kinds of traffic noise:high-high aggregation,low-low aggregation,high-low aggregation and low-high aggregation.High-high aggregation and low-low aggregation exist in large quantities and distribute centrally in cities,while high-low aggregation and low-high aggregation are relatively less distributed in cities.The characteristics of different agglomeration areas are analyzed,and the traffic noise problems under the block level are solved by local improvement of roads and buildings.(2)In the data of the network platform,the more and denser the number of interest points,the higher the vitality value is,and the higher the sound pressure level of traffic noise is.Transportation points of interest(mainly parking lots and bus stations)have the highest positive correlation with traffic noise,and their interpretation ability of traffic noise is as high as 17.72%.Residential points of interest(mainly residential buildings)have the highest negative correlation with traffic noise,and their interpretation ability of traffic noise is as high as 11.09%.The vitality value(representing the relative population density)has a high positive correlation with the traffic noise,and the interpretation ability of the traffic noise is as high as15.28%,indicating that the points of interest and vitality value in the network platform data can effectively explain and predict the spatial distribution of traffic noise.(3)for the space syntax,both axis model and line model,the degree of consolidation and selectivity are high positive correlation with traffic noise,and the explanation ability of integration degree of traffic noise is stronger than the selectivity,due to the integration of the high degree of accessibility is good,that road accessibility and traffic noise that there was a positive correlation between the integration degree is higher,The better the road accessibility,the greater the sound pressure level of traffic noise.By comparing the three index combinations under the two models,it is confirmed that the integration degree of standardized angles and the selection degree of standardized angles under the line segment model have the strongest ability to explain the traffic noise,R~2=0.492,indicating that the spatial syntax index calculated from the road data can effectively explain and predict the spatial distribution of traffic noise.(4)In the traditional econometric model based on multi-source open data,the index combination with the strongest explanatory ability for traffic noise is the integration degree of standardized Angle,points of interest in traffic,points of interest in housing and vitality value,which can be used for a rapid assessment of urban traffic noise.Compared with the traditional econometric model based on urban planning elements(R~2=0.557),it is found that the traditional econometric model based on multi-source open data(R~2=0.566)has higher fitting degree and better explanatory ability,and is more convenient in data acquisition than the urban planning elements.By establishing a geographically weighted regression model,it is found that there is spatial heterogeneity in the explanatory ability of four types of indicators to traffic noise in different regions,and adjusting the indicators with higher influence degree in each region can achieve the purpose of improving traffic noise.This paper forecasts the traffic noise and control the introduction of concept of urban planning,in numerous open data screening effective evaluation index as the research basis,in combination with a variety of research methods and models,explore the open multi-source data for spatial distribution and the influence law of traffic noise,the forecast of traffic noise in innovation and control field has been made,The findings can also guide urban planning.