Research On Meteorological Data Processing And Sharing Platform Based On SOA Architecture

There are many meteorological data sharing platforms at home and abroad,most of them are traditional MVC architectures.Their databases are traditional relational databases such as MySQL and Oracle.MVC architecture and relational database solutions are becoming more and more obvious in terms of software scalability,maintainability,interoperability and storage performance as the business needs and weather data continue to increase.The paper introduces the design and implementation process of SOA-based meteorological data sharing platform in detail,and uses SOA architecture mode under Java platform to realize platform reusability and simplify system development.The MongoDB database is used to store data to accommodate large-volume data queries and read and write operations.At the same time,the PSO-RBF algorithm was introduced on the platform to predict the temperature and construct a temperature forecasting service,as follows:(1)The data comes from the meteorological data of the automatic weather station in Foshan City,Guangdong Province in this paper.The data is organized and built,and the various types of data are standardized and classified coding through the analysis,integration and analysis of the original meteorological data.(2)The paper analyzed the overall design goals and development criteria of the platform according to user needs,the business process and software architecture of the platform,and related design of each functional module and database of the platform.(3)The paper introduces the implementation process of RBF algorithm,and the problem of slow convergence and under-fitting due to improper selection of parameters is used.The optimization of PSO is used to improve the prediction accuracy.The PSO-RBF algorithm is used to train the temperature data.It is predicted that a temperature forecasting service will be built on this basis to enrich the platform function.(4)The paper introduced the original meteorological data analysis format and analysis process,the platform development framework construction process,the specific implementation method of each functional module.The paper introduced the POI framework in the development process,and realized the sharing service of meteorological data by POI.The front-end framework Bootstrap is used to implement the front-end interactive page of the platform.Finally,the function and operating environment of the platform were tested to verify the feasibility of the platform.