Big Data Organization Method And Application Based On Time-space-class Cube

With the advent of the era of big data,the rapid development of the Internet of Things and Internet technology,the construction of smart cities has been significantly accelerated,and big data with temporal and spatial properties has seen explosive growth.How to organize and manage massive multi-source heterogeneous spatiotemporal big data has become a key problem to be solved in urban construction.As the first pilot of smart city construction since the establishment of the Ministry of Natural Resources,Dongying's space-time big data is more consistent with the "unification of the two" responsibilities of the Ministry of Natural Resources.Considering that the current spatiotemporal big data classification system fails to properly link up natural resource data,and also lacks expansibility and sharing.this paper proposes an extensible,sharable and integrated spatiotemporal big data classification system for smart cities based on the aggregation content of spatiotemporal big data in the pilot construction of Smart Dongying.Based space-time data,the system is divided into the basic spatio-temporal data,the public thematic data,the Internet of Things real-time sensing data,the Internet grabs data online,the natural resources and the local characteristic extension data of six categories.There are six categories of spatio-temporal big data,including 18 primary types,65 secondary types and 24 tertiary types.This classification system provides a reference for the collection and aggregation of spatio-temporal big data in the construction of smart citie,so as to avoid problems such as repetition and omission of data collection.In order to solve the problems of non-unique data identification,difficult data organization and incomplete data storage information in the pilot construction of Smart Dongying.this paper proposes a concept model of time-space-class cube based on multi-grid.The main research object of this model is structured spatial data.The idea is to construct a global multi-grid framework,use the spatial index of spatial filling curve Z curve to encode the grid,and then use the minimum bounding rectangle as the index of spatial data.By transforming the data into a unified spatial reference system,the expression of two-dimensional spatial data is transformed into the expression of one-dimensional planar grid,and the unique spatial identification of data is realized by assigning unique spatial codes.In the process of data storage,the grid coding,spatial information,temporal information and attribute information of spatial data are integrated and stored in the database table through the conceptual model,so as to realize the comprehensive integration of spatial location and three dimensional information of spatial data.For the unstructured data,the attribute information of unstructured data is linked to the corresponding spatial data in the way of data linking,which is used as the extension of spatial data,so as to realize the comprehensive organization of heterogeneous spatiotemporal big data.Finally,the multi-level grid framework proposed in this paper is integrated into the smart Dongying spatiotemporal big data platform by using the idea of encapsulation through layers.And the conceptual model is successfully applied to the data organization and storage in the pilot project of smart Dongying construction,which further verifies the practicability of the model.