Research On Division Of Urban Functional Areas Based On Multi-source Attributes

With the development of economy and society,the speed of urbanization in our country is getting faster and faster,the level of urbanization is getting higher and higher.The development of urbanization has entered a critical period,but a series of urban problems accompanying it have become more and more prominent.The division of urban functional areas is of great significance in solving the problems of urbanization and rationally distributing urban spatial structure and resources.The accurate division of urban functional areas,on the one hand,helps to show the unreasonable urban spatial structure in the current urbanization process,and on the other hand helps to provide planning guidance for future urban development.In traditional identification of urban functional areas,methods based on remote sensing image recognition or expert evaluation and field investigation are often used.These methods have the disadvantages of difficult data acquisition,high time and labor costs,poor timeliness and so on.With the development of big data in recent years,the division of urban functional areas based on POI data has become a research hotspot.In this paper,a kernel density estimation method based on multi-source attributes is proposed,which is used to carry out the experiment of division of functional areas in the downtown area of Tianjin,and the precision of the division of functional areas is improved.The main research work of this paper is as follows:(1)The OSM data and POI data are converted into WGS-84 coordinate system to realize multi-source data fusion.In this paper,the OSM road network data and POI data are preprocessed,the OSM road network data are screened and cut,and divided into blocks.The POI data of Baidu Map and Auto Navi map are reweighted,coordinate converted and reclassified.(2)Use the POI quantitative identification method to divide urban functional areas.Taking the central urban area of Tianjin as an example,based on the POI data,the feature vector FD of block and the type ratio CR are constructed,and the functional attributes of the blocks are determined by the type ratio,and the functional area division of the central urban area of Tianjin is completed.Comparing the experimental results with the actual situation in the central urban area of Tianjin on Baidu Maps,the division is roughly accurate.Calculated by the result scoring method,the accuracy rate of urban functional area division based on the POI quantitative identification method is 62.86%.(3)Aiming at the shortcomings of the POI quantitative identification method,this paper proposes a method to divide urban functional areas by using kernel density estimation and adding public awareness attribute and area attribute of POI data as the impact factors,to realize the fusion of multi-source attribute.The division of functional areas in the central city of Tianjin has an accuracy rate of 81.8%,which has a certain improvement compared with the classification of urban functional areas using the POI quantitative identification method.The method of division of urban functional areas based on the kernel density estimation proposed in this paper adds the public awareness attributes and area attributes of POI,more comprehensively considers the influence factors of the POI data that should be divergent and the divergent impact values of different types of POI data are related to the public awareness and area attributes of POI.Through experiments,the accuracy of division of urban functional areasis improved,which has certain reference significance for urban development planning.