Research On Multi-agent-based Quick Query For GIS

With the fast development of the computer technology and the expanding of the social demand,the application of GIS becomes more and more popular. Facing massive and complex spatial data, theproblem that how to improve query efficiency becomes an important task in GIS.This dissertation first introduces spatial data characteristics, and describes spatial data models,organizational strategy and spatial database. This thesis suggests two ways to improve the queryefficiency, one is Agent and other one is spatial database indexing technique. Therefore, setting GISbased air transportation of economy information system as background, the paper made the followingcontributions:1）Improvement of spatial database index. Spatial indexing efficiency is often evaluated byseveral aspects, such as the index’s storage efficiency, query efficiency and update efficiency. Thepaper introduces Hilbert R-tree based on clustering after large number of studies and research andcompares them in capability experimentation. The result is that the new index is better than others,and the new one can improve the efficiency of spatial query system and optimize the systemperformance.2）Using the technology of Agent in GIS. The paper discusses the advantages that Agent cansolve query efficiency problem to which the massive and complex data lead and design anagent-based GIS architecture. GIS agent is categorized as three types: Facilitator, function agent andinterface agent. Secondly, GIS based on multi-agent is implemented on multi-thread technologies. Itcan improve the efficiency.