The Research On Modeling And Storage Optimization Of Meteorological Equipment Data Cube

Meteorological equipment,as the infrastructure of meteorological services,is rich in variety and large in quantity,and it has been widely used in various fields.Meteorological services have benefited all walks of life.A large number of meteorological equipment data resources have been produced in the research,production,and use,which has been increased the management and maintenance cost of meteorological equipment and make a great challenge to the improvement of meteorological equipment.With the popularization and application of big data technology,these huge amounts of data become highly valuable resources and provide decision support for the improvement and management of meteorological equipment.Therefore,the utilization of these data will have an important influence on the research,development and management of new meteorological equipment.This paper starts with the classification of meteorological equipment,combine the attributes of many kinds of meteorological equipment analyzed with the attributes hierarchical inducted.The attribute sets are determined,which concludes basic information,function,technical performance index and the main technology;According to the sets of attributes,the dimensions and dimension hierarchy of meteorological equipment data cube are determined;Meanwhile,the facts of data cube are confirmed based on topics of interest to different types of users;At last,a data cube model for meteorological equipment is constructed,and the storage structure of meteorological equipment data cube is given.After the completion of establishing the data cube model of meteorological equipment,this paper proposes a data cube caching strategy based on analysis of user search behavior,which from the point of view of improving the efficiency of data cube OLAP operation.Firstly,the correlation degree of behavioral habits among users is calculated by the semantic analysis of the users' behavior data,which is reported in the collection,and combine with the user access frequency;Secondly,the correlation degree is used as the threshold condition in the BUC algorithm to compress the volume of the data cube;Finally,an improved cache replacement algorithm(C&T)based on LFU algorithm is designed on the basis of considering correlation degree and time.And Redis is used as cache storage media.The experimental results show that the efficiency of C&T cache algorithm is about 60% higher than that of unused cache and about 20% higher than LFU.In practical application,according to the needs of users,the high availability data cube query application system is designed and implemented.The system uses two machine and cluster mode to achieve high availability on the whole.The design and implementation of data cube creation and data query module are explained in detail.Finally,there are two ways to interact with users: mobile client and desktop client.