Research On Meteorological Data Processing And Visualization Platform Based On Microservices

At present,with the continuous development of information technology,meteorological science has been integrated into the weather forecast and early warning,agriculture,aerospace,environmental protection and other fields,the type and scale of meteorological data is also increasing.However many meteorological data processing and application system at home and abroad are mostly adopt SOA architecture,monomer architecture as the number of meteorological data and access increasing,the software architecture will become more and more bloated,it not only increases the difficulty of development and maintenance,but is not enough to cope with the high concurrency problems caused by the increase in the number of visits to the system.To solve the above problems,this paper designs and constructs a set of meteorological data service platform based on micro services.The services are fully decoupled,and each micro service can be independently developed,deployed and run.At the same time,Redis cache database is introduced to improve the efficiency of database reading and writing.The platform not only meets the user's demand for meteorological data,but also provides a new solution for the sharing and reuse of meteorological data resources.The main work of this paper is as follows:(1)The platform uses Java development language,uses Spring Boot,My Batis and other frameworks to complete the construction of the whole system,through the Spring Cloud Eureka,Ribbon,Hystrix,Feign,Zuul and other components to complete the governance of micro-services,using NCL,D3.js and other technologies to achieve the visualization of different meteorological data formats.(2)Analyzed the development objectives and guidelines of the platform,and designed the overall framework of the platform according to the user needs,and designed each module in detail,as well as the database architecture and table in detail.(3)Using the historical data of meteorological stations for training,an improved Adaboost-C4.5 algorithm prediction model is proposed.PSO algorithm is used to optimize the weight of Adaboost-C4.5 weak classifier to improve the accuracy of rainfall prediction.The Pso-Adaboost-C4.5 algorithm model is integrated into the system to form a weather forecast service,which greatly improves the practicability of the meteorological data service platform.(4)Platform using IO streams and a variety of interface class implements the data of meteorological sites and nc format of satellite data parsing library,in addition,the platform implements including the basic service,data sharing,data visualization and multiple modules such as weather forecast,and through the test tool of each function module was tested,verified the system design to meet the requirements set.