| With the development of wireless communication and Internet of things,the de-mand for location aware and location-based services is increasing.The technology re-search based on wireless location is deepening.Now,GPS and Beidou positioning sys-tem already provide high-precision outdoor positioning and navigation service to people through satellite signals.However,the satellite signal strength drops sharply as it passes through the reinforced concrete structure,rendering it unable to provide complete and valid positioning information.Therefore,GPS and Beidou positioning system are not suitable for indoor scenes.At present,there is no indoor positioning technology that is applicable to various indoor scenes and effective.Low-cost and high-precision indoor positioning technology has been a hot research topic.Inspired by the biogenetic geomagnetic field in the biology such as the messenger pi-geons and the lobster,people have gradually discovered and studied a technique of using the earth's magnetic field for indoor positioning.Geomagnetic indoor positioning tech-nology does not depend on any infrastructure equipment,even no regional restrictions.With the advantages of low cost and high accuracy,it gradually becomes an emerging technology which has attracted much attention.Based on the research of indoor magnetic environment and geomagnetic indoor positioning technology,we firstly designed a geo-magnetic signal acquisition device based on Android mobile platform,which can simply and conveniently realize indoor geomagnetic signal perception and acquisition.Using the geomagnetic signal acquisition device,the feasibility of the geomagnetic field in the indoor scene as a positioning carrier was verified from two perspectives of stability and discrimi-nation,respectively.Kriging was subsequently selected by comparing the advantages and disadvantages of the five common spatial interpolation algorithms.Interpolation algo-rithm will be collected geomagnetic data to form a 0.1m geomagnetic database,and draw the indoor geomagnetic reference map.Secondly,by analyzing the existing common geo-magnetic matching algorithm,we know that the particle filter algorithm has the problem of particle degradation.Although some researchers reduced the phenomenon of parti-cle degradation by introducing the method of resampling,this method will also bring about the problem of insufficient particle diversity.In this dissertation,the unobtrusive Kalman filter algorithm is combined with the RTS smoothing algorithm to generate the importance probability density function.The MH sampling algorithm in the MCMC algorithm is used to optimize the processing of the sampled particles,resulting in an improved RUPF-MH algorithm.The algorithm can reduce "particle degradation" and improve the diversity of particles so that it can better adapt to the needs of geomagnetic indoor positioning scenes.Finally,this thesis designed a geomagnetic indoor positioning system based on Android smart phone,and carried out a detailed design analysis of the modules contained in the positioning user client and background processing server respec-tively.In order to improve the convergence speed of the geomagnetic matching algorithm,a two-dimensional HV geomagnetic fingerprint storage and matching model is selected in the positioning system.For the sake of improving positioning accuracy,track prediction technology is introduced into positioning system,a time interval detection and dynamic threshold are proposed in gait detection.Through the experimental analysis,the feasibility and rationality of the design of the whole geomagnetic indoor positioning system are verified.The system has an average precision of 2 meters,which can provide daily indoor positioning service for users.It also verifies the effectiveness and superiority of the two-dimensional HV geomagnetic matching model and the improvement of particle filter algorithm in the geomagnetic localization scenario. |
PDF Full Text Request