Spatial Data Matching Method Of City Underground Pipelines Considering Semantic Similarity

Underground pipeline is the basic tunnel for city material flow,energy flow and information flow and also the important infrastructure and lifeline of the city.Influenced by pipeline managing mode and different application of pipeline spatial data,there are two types of pipeline geo-information system in current information of the city:the integrated pipeline GIS which use the integrated pipeline as data source and serve mainly for the urban planning;the professional pipelines GIS,which use the professional pipelines as data source and serve for the operation and management of the pipelines authority sector.Because of influenced by feature classify of pipeline information,standard of pipeline scheme,differences of collection and information platforms,the two types of pipeline applications have formed two resources of spatial data where differences in semantic,data model and data precious,though facing to the same pipeline in the same district.So,the integration,combination,sharing and exchanging of different type of pipeline data have gradually become the serious problem of the undertaking applications of city pipeline information and future lead to let the underground pipeline information systems in city constricting become the information isolated islands.So,comparing,sorting the spatial location,attribute information of the two data,implementing the matching and combination of pipeline data have inevitable choice for improving the repetitive detection for urban pipeline data,reducing production costs of the integrated pipeline and improving the professional quality of the data pipeline.This paper proposes a underground pipeline semantic similarity calculation model(Pipeline Semantic Similarity),and it studies the pipeline spatial data matching method on the basis of model and the semantic and topological feature.The research contents and results in this paper are as follows:Firstly,according the semantic feature of pipeline data,based on the traditional semantic similarity calculation model of MD3(Triple Matching-Distance),combined the concept name,semantic distance,attributes and spatial characteristics and other factors of pipeline data,this paper proposes a underground pipeline semantic similarity calculation model and this model solves the problem of semantic inconsistency in underground pipeline spatial data matching process.Secondly,referencing the traditional data matching method of spatial vector data and combined with the topological feature of pipeline data,this paper re-constructs pipeline data structure from 'point-line'into 'node-curve',uses the P-MD model to match pipeline data considering its spatial and semantic feature,and takes the gas pipeline data as a matching test example.The result shows that the P-MD model effectively solves the problem of matching synonym pipeline points in the local region.