Research On Meteorological Data Processing And Visualization Platform Based On Distributed Architecture

At present,with the development of agriculture,fishery and animal husbandry,aerospace industry,as well as the increase of the demand for early warning of natural disasters and the treatment of environmental pollution,the importance of meteorological data is increasingly prominent.The application scenarios of meteorological data are also diversified and multidimensional.Although there are many meteorological data query application systems at home and abroad.But they mainly adopt the traditional MVC three-tier architecture.With the increasing variety of meteorological data,the increasing amount of data and the increasing number of concurrent visitors,the traditional software architecture is not enough to deal with the system problems caused by large amount of data and high concurrency.In this paper,a meteorological data application platform based on distributed architecture is proposed.Its service modules are independent of each other,different modules communicate and call through lightweight protocols(HTTP,rest,etc.)to achieve division of labor and cooperation,then the integrity and reliability of the system application are guaranteed.Compared with the traditional weather information system,this software architecture has lower coupling and higher flexibility.In addition,redis data caching technology,database optimization technology and load balancing technology are introduced to further optimize the software architecture of the system platform;so that it can have certain high concurrent load capacity,large data processing capacity and fast response speed.The performance and pressure tests show that the system platform has good stability and reliability.This also proves the effectiveness of the distributed architecture and various optimization methods.In the application of meteorological data,this paper introduces the process of classification,processing,analysis and storage of the original meteorological data in detail.In the background system,JAVA development language,NCL,D3.JS and other visualization technologies are used to perform multi-dimensional,multi-attribute and interactive visualization processing on conventional meteorological data and Net CDF file data.Using POI and other tools to achieve the integration and sharing of meteorological data.In addition,a decision tree algorithm is used to train meteorological data,and the original algorithm execution flow is improved according to the application scene of the data stream,At the same time,in order to find the optimal solution of the parameters in the dynamic system,PSO-ds CART algorithm model is finally put forward to predict the precipitation and temperature of meteorological data,and it is integrated into the system platform to provide meteorological data prediction service for users,further improving the practical application value of the system platform.