Design And Implementation Of Personnel Positioning System Based On LBS

With the development of mobile Internet and GIS information technology, the positioningsystem based on LBS (Location Based Service) is used more and more widely. Meanwhile, thelocation information datas accumulated by the various systems also has geometric progressiongrowth; how to organize, manage and express these vast amount of spatial data effectively andhow to make full use of these spatial data, excavate the hidden knowledge and rule, and thenprovide data support for decision making, so as to enhance the intelligent level of the positionsystem, have become important research contents for spatio-temporal data model and spatial datamining technology. Therefore, this paper aims to fulfill the requirement of data mining andanalysis in personnel positioning system based on LBS, emphatically studies the apatio-temporaldata model and the cluster analysis algorithms in spatial data mining.This paper first studies OPH spatio-temporal data model, DBSCAN clustering algorithmand K-means clustering algorithm, pointing out their existing problems. And then, aiming tooptimizing and simplifying the OPH spatio-temporal data model, the OH spatio-temporal modelfor personnel positioning data is proposed. Its structure is simple, more suitable for person objectand describes the spatio-temporal feature of personnel positioning data more completely. Theretrieving experiment on the model justified its feasibility, thus lay the foundation of further datamining. Furthermore, the paper deeply analyzes DBSCAN and K-means algorithm, based onwhich an improved algorithm is proposed, enabling retrieval of outlier points and detectingoutlier points. The results show that the improved algorithm is more precise on outlier pointsdetecting and the effect of clustering is better, the paper also propose an algorithm for comparingperson trajectory similarity, solving the problem of similarity comparing between point setsdirectly. The experiment proves that the algorithm is feasible on person trajectory similarity.In the end, based on the theory and methodology the paper present, the paper design andimplement a personnel positioning system based on LBS for the realistic project requirement.The system is able to real-time position searching, history position searching, spatio-temporal searching, space searching, outlier area analysis and person trajectory similarity allocationanalysis.