Modeling Research Of Spatial Data Based On Object-Relational Model

The spatial data has huge data quantity and visibility, how to carry on effective manage ment and analysis to the spatial data is a hot research topic GIS at present domain. GIS is a kind of mature tool and technology that operates spatial information now. However, it can only implement the limited operation to the spatial objects, cannot carry on transformation, index, interlink age and excavation to the spatial data, and hard integrate other information with combining map variable to set up and analysis new variable. Meanwhile, relational model in the traditional DBMS and the object- oriented modeling do not completely suit the spatial application domain. The relational model can handle topological relation well, but shows inabilities to that complicated level crossing the spatial region. Object-oriented model using software method based on the principle of the customer defined data type supports the abstract data type, especially spatial abstract data type, and possesses polymorphism and inheritances, and it can deal with topological and the layer relation, but hardly processes important continuous phenomenon in space. Therefore, it is necessary that construct the highly effective spatial database system to realize effective processing to spatial data, and the most important thing is establishing suitable spatial data model for this. This thesis precisely embarks from this demand, uses the method of integrating object-oriented design principle and the mature relational data model, studies several questions of spatial data model from establishment to realization on the object-oriented relational technology, and expects providing some references for processing complex space data conveniently.The spatial data type is complex and diverse, and divided into spatial vector data, spatial raster data, spatial topology data, and spatial network data according to the difference that the spatial data are organized and expressed in the computer. Nowadays mostly research concentrates on spatial vector data model to the spatial data model, and a few spatial topology data are involved. This paper achieves systematic and synthetic discussion for the four spatial data modeling of kind types and enriches research content of spatial data model.The traditional description tool of conceptual model such as ER model, UML, is insufficient to describe the spatial data. We can able to express the spatial inferential semantics by increasing some kind of structure to carry on the expansion. The paper introduces pictogram expansion model which is precisely based on this kind of thought causes the spatial application the conceptual modeling simpler and direct-viewing. Two kinds of spatial information model--field model and object model,which based on spatial application angle, nearly summarize all spatial entity and the spatial structure,and they upholstery solid application foundation for following spatial data model researcch through the research of the two kind of models and the model realizations.Under the object and relation pattern, the article has systematically studied the definition, establishment, expression, memory structure of spatial data, establishment of spatial index, model realization and so on about the four kinds of type spatial data model. Meanwhile, model is validating about realizable and operational by examples, which combine the application in spatial database in the paper.Synthetic study of spatial data model using object-relational technology from first floor, and as well as the several questions mentioned above is the innovation spot of this paper.