Research On Indoor Plane Positioning And Floor Positioning Technology Based On Multi-sensor Fusion

There are more and more indoor scenes that need to be positioned in real life,and user's location-based service has become a hot research topic.For modern buildings,it is equally important to accurately locate the user's plane position and vertical position.With the rapid development of technology and the popularity of smart phones,the application of built-in sensors in smart phones provides new ideas for indoor positioning.Aiming at the crowded and mixed indoor environment,this dissertation focus on the problem of user's plane positioning and floor positioning with the help of wireless local area network technology and built-in sensors of smart devices,and the following aspects of the content are mainly studied.Firstly,a regional subdivision plane localization method based on Wi-Fi and magnetic field fusion is proposed.The method is based on fingerprint matching positioning technology.The main task of the offline phase is to establish Wi-Fi and magnetic field fingerprints,and subdivide the positioning area by clustering Wi-Fi fingerprint.During the online phase,the Wi-Fi-based fingerprint matching algorithm is firstly used to narrow the positioning area,and then the high-precision magnetic field fingerprint matching algorithm is used to obtain the final position of the user.Due to the complex indoor environment,in order to further improve the positioning accuracy,the Wi-Fi data and the magnetic field are separately filtered.Secondly,a multi-floor indoor positioning method based on Wi-Fi and barometer fusion is proposed.The main task of the offline phase of the method is to establish the Wi-Fi fingerprint of each floor.Considering that the air pressure value is relatively stable and has a clear discrimination during a short time,the method clusters Wi-Fi fingerprint with the help of the air pressure value,and divides the Wi-Fi fingerprint of the same floor into the same cluster,and the barometric pressure value is uses to calculate the average floor space.During the online phase,the initial floor and its height are firstly estimated by Wi-Fi data and barometric pressure,and then the user is continuously positioned by the barometric pressure value,and the accuracy of floor positioning is further improved by smoothing and filtering the Wi-Fi data.Finally,the designed experiment is used to verify the plane positioning method and multi-floor positioning method proposed in this dissertation,and evaluates the performance of the experimental results in detail.