Research On Spatial-temporal Model And Fast Retrieval Method For Big AIS Data

Since the 21st century,Automatic Identification System as a compulsory communication technology for intercoastal communication,has provided maritime and regulatory authorities with the most complete,most widely used,most standardized and globally common data of the waterway transportation industry.The use and analysis of AIS data provide reliable data guarantee for safe navigation and collision avoidance of ships,provide information support for decision-making and law enforcement of maritime departments,and lay a foundation for more efficient and fast shipping logistics.However,with the continuous development of maritime economy and the increasing variety of ships,the AIS data gradually presents the characteristics of big data,which brings great difficulties for its management and query.This is especially true in the current era of Internet and mobile computing.In order to solve this problem,researchers at home and abroad have proposed a series of effective methods.However,most of these methods are aimed at a certain aspect of the spatial and temporal characteristics of AIS data.The lack of comprehensive consideration of this characteristic brings difficulties for subsequent management and retrieval.Therefore,based on the storage management of AIS big data and the practical needs of application services,this paper proposes a spatial-temporal data model of AIS data with comprehensive consideration of the spatial-temporal characteristics and an efficient storage and retrieval method to solve the difficult problems of AIS big data management and query.Firstly,this paper summarizes the existing AIS data storage management methods and index technologies through the analysis of the research status at home and abroad,and then analyzes their shortcomings,and proposes the main research contents of this paper.Secondly,based on the analysis of spatial and temporal characteristics of AIS data,a conceptual data storage model with ship as the spatial and temporal object is proposed.Further combining with the distributed environment,a multidimensional relational storage model is proposed to achieve efficient organization and management of mass AIS data.Thirdly,on the basis of the aforementioned spatiotemporal data model,the AIS big data management architecture is proposed.Through the analysis of the writing mechanism of ES,the improved My SQL paging retrieval method and the performance score SUR(Segment Union Remark)optimization strategy based on the writing mechanism were proposed,which realized the efficient storage of AIS data.Based on the multi-dimensional analysis storage model,a "wide table" storage method and an "application-end relational" storage method were proposed to realize the hierarchical storage of AIS data and its navigation conditions(navigation environment,navigation area under water area jurisdiction).Finally,a spatiotemporal data partitioning strategy based on Spatiotemporal-Slice Cube(SSC)is proposed through the segmentation of time dimension and space dimension.On this basis,the Geo Cub R tree is constructed by combining with the Geo Time coding method.Based on Lucene’s index construction method and distributed storage principle,a distributed index framework of invert-geocubr tree was proposed to realize fast retrieval of massive AIS data.In addition,this paper designed the retrieval and experimental schemes,built the AIS data storage management system based on B/S architecture,and verified the efficiency of the writing,storage and indexing methods proposed in this paper.Experimental results show that the storage method and index structure designed in this paper can better cope with complex spatial-temporal relational queries,and provide effective management and fast retrieval services for mass AIS data.