Research On Indoor 3D Positioning Algorithm Based On Map Matching And Multi-source Data Fusion

With the rapid development of urbanization and the rapid popularization of mobile portable devices,the time people spend indoors has increased significantly,and the demand for high-precision indoor positioning services has also increased year by year.Due to the limited accuracy of the original positioning coordinates provided by the indoor positioning system,there is a problem that the positioning error of some track positioning points is large.Therefore,it is necessary to use map matching technology to correct the positioning solution point error and improve the positioning accuracy of the entire positioning system.To provide high reliability,high stability and high-performance indoor positioning services,the indoor 3D map matching algorithm and 3D positioning technology are studied under the background of certain errors in the indoor positioning system solution points.The specific research process and research contents are as follows:(1)Map matching algorithm based on WiFi RSSI time series classifier optimization.In view of the problems of existing map matching algorithms in indoor environment,such as large algorithm complexity,lack of historical context information analysis of positioning data and failure to integrate topology semantic information,a map matching algorithm based on improved HMM is proposed,which uses speed constraints to screen reasonable candidate matching sections,so as to improve the efficiency of the algorithm;Using the geometric information and topological information of the track location points to match the map,the problem of large error location points in the track is solved.Secondly,a GRU based WiFi RSSI timing data path classification algorithm is proposed.The indoor user path classification is carried out through the WiFi RSSI data collected by the user terminal equipment in real time.The path classification results are used to solve the problem that the traditional map matching algorithm based on HMM only considers the adjacent track location points.Finally,an indoor 3D navigation map construction scheme based on semantic topology is proposed,which can provide geometric,topological and semantic information to optimize the accuracy of map matching algorithm.(2)Multi-source optimization-oriented 3D map matching algorithm and localization algorithm.Aiming at the problem that traditional indoor positioning systems and map matching algorithms have not yet solved the problem of 3D positioning,a relative altimetry algorithm based on air pressure self-calibration and fusion filtering is proposed.The defect of the need to place the reference air pressure acquisition equipment floor by floor;the use of the Kalman filter-based algorithm to fuse the acceleration information to improve the vertical height positioning accuracy.Secondly,the 3D expansion based on the 2D map matching algorithm solves the defect that the traditional map matching algorithm is not suitable for the real-time positioning of 3D floor switching and can provide good matching performance in the indoor 3D map.(3)The performance of the above algorithm is evaluated through experiments.The experimental results show that the above algorithm is suitable for indoor 3D positioning scenarios.Compared with the traditional map matching algorithm and positioning method,the matching stability and positioning accuracy are further improved under the background of positioning errors in the indoor real-time positioning system.