Research On Navigation And Positioning Algorithm Based On Wlan Indoor Modeling

In recent years,with the rapid development of wireless networks and artificial intelligence technologies,people have paid more and more attention to location-based services.The global positioning system cannot perform normal indoor positioning because satellite signal is easily blocked by indoor walls.The existing multiple indoor positioning technologies are restricted by positioning accuracy,cost,hardware and other factors,but have not been widely used.The rapid deployment and popularization of wireless local area network(WLAN)provide a good environment for development,application and promotion of WLAN indoor positioning technology.This article focuses on in-depth research on location fingerprint positioning technology in WLAN indoor positioning technology.The main research contents are as follows:First,in view of the low timeliness of the fingerprint library in the WLAN location fingerprint indoor positioning technology,the mean function of RSS,Gaussian process regression model and geographically weighted regression model are studied,and a fingerprint database update algorithm combining Gaussian process regression(GPR)and geographically weighted regression(GWR)models is proposed.The algorithm combines the wireless signal propagation attenuation model with the GWR model to construct the received signal strength(RSS)mean function,which is used to estimate the mean value of the RSS at the fingerprint point to be updated,and the GPR model is used to estimate the residual error of the RSS at the fingerprint point to be updated.Finally,the sum of the mean value of the RSS and the residual error was used as the estimated value of the RSS at the fingerprint point to be updated to realize the adaptive real-time update of the fingerprint database.Second,to solve the problem of low matching accuracy of sorting clustering positioning algorithm in WLAN location fingerprint indoor positioning technology,a matching optimization and distance assisted Wi-Fi positioning algorithm is proposed.According to the user’s position,distance and step size,a class matching deviation detection model is designed to judge the user’s position anomaly and matching deviation;the adjacent elements in the sorted received signal strength(RSS)vector are compared with the set threshold to determine the change position of the sorting feature vector of the point to be located,and the exchange correction is performed to obtain the corrected and merged class matching result;according to the distance between the user’s position determined in the time period m before the positioning and the fingerprint point in the matching class,the abnormal fingerprint points used for position calculation are eliminated,so as to achieve more accurate indoor positioning.Finally,simulation verification shows that the average estimation accuracy of the RSS of the fingerprint database update algorithm based on the fusion of GPR and GWR models proposed in this paper can reach 2.63 d Bm.In addition,the proposed Wi-Fi location algorithm based on matching optimization and distance assistance improves the class matching accuracy by 17% and the average positioning accuracy by 22% compared with the sorting clustering location algorithm.