Research On Indoor Localization Technology Based On WiFi And PDR

In recent years,with the needs of positioning and navigation,rescue missions and the popularity of Internet of Things technology and smart mobile devices,the demand for positioning services in indoor environments is increasing rapidly.Therefore,the research on positioning and navigation technology for indoor environment has attracted extensive attention from researchers.The current mainstream indoor positioning technologies include WiFi positioning technology and Pedestrian Dead Reckoning(PDR)technology.WiFi positioning is a technology that completes positioning by receiving and processing WiFi signals in the indoor environment.It has the advantages of low deployment cost,wide application range,and can provide global positioning.However,the disadvantages are low positioning accuracy,low signal sampling frequency,etc.PDR positioning technology realizes positioning by collecting and processing the inertial sensor data from the device itself,which is a kind of relative positioning technology.The advantages of PDR positioning are high positioning frequency,high short-term positioning accuracy,and the ability to estimate the direction of movement,etc.The disadvantage is that when performing long-term positioning,the PDR system will generate a large cumulative error.Therefore,in order to achieve indoor positioning with high positioning accuracy,high real-time performance and long-term operation in the indoor environment,this paper proposes an indoor positioning system based on the fusion of WiFi and PDR.Its main work includes the following aspects:1.Aiming at the problem of time-consuming,labor-intensive and high cost of fingerprint map construction in the offline phase of WiFi fingerprint positioning,a WiFi fingerprint map construction method based on Gaussian process,GPFM,is developed.By analyzing the construction process of WiFi fingerprint map based on Received Signal Strength(RSS),the defects of the logarithmic attenuation model of RSS signal were deeply excavated,the RSS signal distribution fitting model was improved,and a Gaussian process-based GPFM signal fitting model was proposed.The experimental results show that,compared with the traditional method of collecting offline fingerprints called site survey,the proposed WiFi fingerprint map construction method can provide high-quality offline fingerprint maps while significantly reducing the amount of offline data collection.2.Aiming at the problems of long matching time and large positioning error in the online stage of WiFi fingerprint positioning,a WiFi fingerprint positioning algorithm(Fingerprint Clustering and Fusion Distance-based Localization Algorithm,FCFDL)based on fingerprint clustering and hybrid distance is proposed.By analyzing the characteristics of RSS signals,the reasons for the large positioning errors in the traditional fingerprint positioning process are deeply excavated,and the fingerprint clustering and fingerprint positioning methods are improved.The experimental results show that the proposed WiFi fingerprint positioning method can provide higher positioning accuracy while reducing the computational cost and improving the computational efficiency.3.Aiming at the problem of low sampling frequency of WiFi signal in practical applications and the requirements of positioning accuracy,continuity and real-time performance of WiFi/PDR fusion positioning system,a WiFi/PDR fusion positioning method based on particle filter(PF-WiFi PDR)is proposed.By analyzing the advantages,disadvantages and complementarity of WiFi positioning and PDR positioning,PDR positioning based on Step-and-Heading System is improved,and a fusion particle filter positioning method based on WiFi and PDR is proposed.The experimental results show that the proposed WiFi/PDR fusion positioning method can make full use of the complementarity of WiFi fingerprint positioning and PDR positioning,and effectively solve the problems of trajectory jump in WiFi positioning and cumulative error in PDR positioning.