| Indoor positioning technology has always been a research focus in the field of navigation and positioning.The success of many Internet of Things(IOT)applications relies on indoor positioning that is expected to provide reliable position information.Among the numerous indoor positioning methods,since multiple Wi Fi access points(AP)can be detected indoors and their signals are easy to measure,the fingerprint positioning method based on the received signal strength of Wi Fi has become one of the most popular positioning technologies.However,although the densely laid AP devises improve the information dimension of positioning matching,the appearance of unstable AP signals not only leads to the bloated fingerprint database,but even reduces the accuracy of positioning.Therefore,with the goal of optimizing the structure of the fingerprint database and improving the positioning accuracy,this paper carries out research from the offline database construction stage and online matching stage of fingerprint positioning,and strives to reduce the amount of data storage and improve the positioning results.The main work and innovations are follows:(1)Considering the problem that traditional clustering method is difficult to divide physical space effectively,and positioning error is large due to the signal source is unstable,this paper proposed a simplified AP matching location algorithm based on fuzzy clustering.According to the proposed algorithm,in the offline stage,target space of large area is divided into multiple partitions by the characteristics of the signal source,comprehensively considers the stability、visibility and redundancy characteristics of the signal source in each partition,establishes the smallest AP identification set in the area,overcomes positioning fluctuations caused by unstable AP signals while reducing fingerprint comparison time.In the online stage,the traditional Euclidean distance is improved,and the weights of neighboring points are assigned based on the stability of the regional AP,and the speed constraint between adjacent moments of the user to be located is used to filter the positioning outliers,to overcome the changes in the environment and signal sources.The unfavorable influence of the locating error is reduced.Tested on actual scenes,the proposed algorithm has an average positioning error of no more than 0.977 m while reducing the amount of data storage.Compared with the existing classic positioning methods,its positioning accuracy is improved by more than 15%.(2)Considering the problem that the reference points(RP)are densely distributed in the positioning scene,the heavy task of constructing fine-grained fingerprint database and the large positioning error caused by the signal fluctuation of AP,this paper proposed a fingerprint database construction algorithm under the condition of dual scale AP selection.Based on the high dimensional characteristics of local sparsity and common neighbor similarity of the received signal,the algorithm completes the division of the positioning area,takes into account the discrimination and visibility of the AP signal,provides a more efficient and reasonable AP set for online computing,and interpolates to establish a virtual fingerprint database to reduce the workload of offline collection.In the online stage,the problem of poor judgment of the area of edge points to be located is overcome through secondary fuzzy matching,the matching scheme of selecting virtual fingerprint database is optimized.The positioning results are detected twice in the way of density screening to avoid the positioning error caused by outlier nearest neighbor points.Tested on actual scenes,in the three RP density scenarios of 1m,1.6m and 2.2m,the proposed algorithm’s positioning accuracy is improved by more than 6%,13% and 18%,respectively,compared with the classic algorithm.(3)In order to better apply the research topic into real life scenarios,this paper designs and develops a indoor positioning system based on We Chat applet.Taking the daily positioning needs of users as the background,the system aims to achieve convenient positioning and information exchange,and uses applet,cloud server and web page as the main carrier to provide digital convenience for system users in different roles.The system uses We Chat applet to realize the front-end interaction of indoor positioning and information collection,and uses the super computing power of cloud server to realize the positioning algorithm logic proposed in the paper,and at the same time provides a unified display of graphical interface for positioning service providers.By transplanting the realization of the positioning function to the application programs commonly used by users,the system has carried out technical innovation on the original design mode,and further improved the user experience. |
PDF Full Text Request