Research On Population Spatialization Method Based On Individual Data Deduction

Human beings are the main body of social and economic activities and the basic elements and driving forces of social and economic development.As the main indicator reflecting economic and social activities,population data is widely used in the research of various aspects of social and economic development.The census data is the authoritative source of population information that is often used in current research,and provides accurate data support for scientific analysis.However,in practical applications,using the statistical summary data of census directly has poor timeliness and low spatial resolution.It also has the problem that the administrative unit is inconsistent with the natural unit.The in-depth development of census data and the construction of a multi-scale,high-precision population spatial database are of great significance for population-based research.In order to solve the above problems,scholars based on the census data and the population distribution indicator,using a variety of models and methods and the demographic data based on administrative divisions combined with natural unit data to achieve population spatialization and construct grid population database.The population spatialization model is increasingly rich,and the accuracy of the results is continuously improved.However,due to the complexity of population distribution law and the limitations of existing modeling methods,the current research results mainly focus on the spatialization of population(or population density).There are few studies on the spatialization of population attribute data.In order to cope with the new demand for census data and the spatialization method of innovative population attribute in scientific research and social application,this paper combines the two researches of synthetic population microdata generation and population spatialization,and conducts population spatialization method based on individual population data deduction.Research,build urban artificial population database and multi-scale population and attribute grid database.The logical ideas of this paper are as follows:Firstly,the literature review and theoretical research on population spatialization model and synthetic population generation algorithm,which lays a theoretical and methodological basis for this paper;secondly,the iterative proportional updating algorithm in synthetic population generation is applied to Chengdu five urban districts of total 92 streets,two micro-databases containing 1,141,413 households and 3,411,592 individuals were generated and analyzed separately.Finally,based on the generated Chengdu micro-database,the area weight model is used to generate two different scales of population grid datasets of 1km×1km and 500m×500m,and the empirical analysis of population attribute spatial characteristics is taken as an example.The main conclusions of this paper are:(1)Based on population micro-sampling data and census data,the iterative proportional updating algorithm in synthetic population generation can be used to generate micro-databases for families and individuals in China.This method has short calculation time and high fitting precision.And it is an effective method for in-depth development of census data,construction of high-precision micropopulation data sets,and removal of population privacy information.(2)Based on the artificial population micro-database,the population spatialization method based on the area weight model is simple and effective,and can effectively construct a multi-scale population grid database,especially to generate a small-scale population grid database.At the same time,this database not only contains the quantity information of the population,but also contains some attribute information of the population,which provides rich data support for conducting high-precision research related to population.(3)The grid data of different scales have their own advantages and disadvantages.For the grids of 1km×1km and 500m×500m used in this paper,the grid of 500m×500m is more elaborate.The characterization of the boundary of the original area is more detailed,and at the same time,the information inside the area can be more completely preserved;the advantage of the grid of 1km×1km is that the calculation amount is small,and the units inside multiple areas can be integrated,at the grid scale.The difference between the units can be reduced,so that the changes between the grids are relatively flat.