Design And Implementation Of Analysis And Visualization System Of Industrial Space Pattern Based On HTML5

At present,the visualization of industrial spatial pattern is mainly dependent on the traditional geospatial analysis tools,and its analysis results are saved as static visual results.The key lies in the hardware,the researcher's level of professionalism and the workspace.And the visualization of research results often just in a single template,namely static visualization method.It is not suitable for the current efficient and flexible research and the visualization results are difficult to share.At the same time,the analysis model applied to the spatial pattern in the traditional geospatial analysis tool is too basic and difficult to be used in the analysis of thematic research.For example,the Geographically Weighted Regression(GWR)analysis does not consider the time factors.The subsequent researchers have improved the time and space geographic weighted regression model(GTWR)by adding time factor,but it is still prone to fall into the local optimum when modeling and analyzing.In view of the above problems,the following two main tasks are carried out:(1)In order to solve the problem that the selection efficiency of characteristic variables in GTWR modeling is low efficiency and easy to fall into the local optimal problem.We combine the Akaike information rule(AIC),and let AIC value difference 3 as the basis to improve the GTWR variable selection method.The key principle of the improved method is: first,when initializing,the two feature variables are pre-selected into the selected variable set to improve the method efficiency.Then,according to the difference of the AIC value difference,the variable is selected from the pre-selected feature variable set to the selected variable set.Finally,variables deleted from the selected variable set on the basis of removing criteria.The results of experiment show that the proposed method significantly improves the efficiency of feature selection.(2)To solve the problem that the current industrial spatial pattern analysis tools are strong dependence on hardware,the operators and have difficult in visual the analysis results,the HTML5 based industrial spatial pattern analysis and visualization tools have been realized.First,the framework,database,system function and theme style have been designed,and then the development environment and visualization methods have been set.Secondly,the concrete realization of the system is finished based on HTML5,Web GIS,Web GL and Java Script and Python programming language.Users can browse,query and search the results of the spatial pattern of the industry directly on the web page.Our system enrich the expression form of industry spatial distribution,break the tradition of research in the fixed workspace and operated by professional.