Research On Method Of Building Meta Semantic Object Spatial Temporal Data Model Based On Hadoop

With the development and popularization of mobile devices and Internet of things technology,the scale of spatio-temporal data is expanding rapidly.Therefore,how to express and manage massive,multi-source and heterogeneous geo spatial information is urgent.The conventional method of geo spatial data organization usually express the complex geographic entities through point,ployline,polygon and meta.and easily lose the correlation information between attributes and semantics of geographic entities.At the same time,GIS mostly uses a single machine relational database system to store spatio-temporal data,which has weak scalability and low reliability.To solve the above two problems,this paper will constitute the geographic space reduced to the minimal logic unit---Mete Semantic Object,and combined with the Hadoop framework and column store database,MapReduce parallel computing module,and design the Meta Semantic Object Spatial Temporal data Model based on Hadoop(HSST_Model),and study the distributed storage and retrieval methods of spatiotemporal data.The main technical contents include the following aspects:(1)The method of objectification expression and encapsulation of geospatial entity information.Based on the open source cloud platform Hadoop,and combined with object-oriented methods and semantic constraint models constitute a geographical entity element decomposed into Mete Semantic Object,Combined Semantic Object and Aggregation Semantic Object,describe the attribute sets,geometric features,method sets and semantic relations of all kinds of semantic objects,and provide theoretical support for the organization of geographic spatio-temporal data.(2)The method of building Meta Semantic Object Spatial Temporal data Model based on Hadoop.This paper discuss the space relationship calculation method of different mete semantic objects,including element metric relations,topological relations,meta semantic object as the smallest organization of spatio-temporal information unit,construction of Meta Semantic Object Spatial Temporal data Model based on Hadoop,The temporal,spatial and attribute information of meta semantic objects and spatio-temporal objects are expressed by the model.(3)The organization method of geographical spatiotemporal data under Hadoop architecture.By analyzing the characteristics of spatio-temporal data,and based on the distributed storage framework of geographical spatio-temporal information,We design the physical storage structure of the HBase database with the meta-semantic object as the minimum storage unit,elaborate on the time-space object access method,and construct a spatio-temporal index structure to improve the query efficiency of spatio-temporal data.(4)The application of the model to the organization and query of traffic spatiotemporal data.According to the characteristics of traffic spatio-temporal information,the HSST_Model is used to organize and store the traffic spatio-temporal data.Finally,the data storage and retrieval efficiency of VCP-Tree index is verified by experiments.This paper through building the Meta Semantic Object Spatial Temporal data Model based on Hadoop,and studying spatio temporal information distributed storage and retrieval method and on this basis,It has realized the effective organization and management of large-scale geo spatial data,and provided a new solution for the increasingly complex geographic spatio temporal information in the process of computer storage.