Design And Implementation Of Indoor Positioning System Based On Smartphone Sensors

With the increasing demand for location-based services,the application prospects and commercial value of indoor positioning technology is increasing.The indoor positioning technology based on smartphone sensors is developing rapidly and has been widely used.Among them,high-precision,multipath interference-based visible light positioning(VLP)technology based on mobile phone image sensors and low-power,inertial sensor-based pedestrian dead reckoning(PDR)without external signals Technology has attracted much attention in indoor positioning research.This thesis aims at the problem of reduced VLP accuracy and positioning discontinuity based on image sensors under pedestrian walking,and the problem of PDR technology's accumulation of positioning accuracy over time without correction of external position information.A combined indoor positioning system based on PDR and visible light positioning are proposed.the attitude estimation module provides the attitude information for the VLP based on the image sensor,and combines the LED geometric information and ID information to realize the absolute positioning of the pedestrian.In ID identification and tracking,this thesis proposes an ID identification and tracking algorithm based on finite state machine,to avoid repeated identification of the same lamp.At the same time,the attitude estimation module provides pedestrian heading information for the PDR module,and combined with the acceleration data,the PDR module realizes the relative positioning of pedestrians.In the PDR module,this thesis designs an adaptive threshold step detection algorithm,which improves the robustness of step detection.A particle filter is used to fuse the VLP result and the PDR positioning result.When the visible light positioning result cannot be obtained,only the PDR output result is used,and the fusion positioning result is output when both results are available.The experimental results show that in the experimental scenario where 4 LED lights are arranged in the room area of 5m×5m×2.98 m,the combined indoor positioning system based on mobile phone image sensor and 9-axis sensors(gyroscope,magnetometer,accelerometer)proposed in this thesis effectively improves the positioning accuracy,and the average dynamic positioning accuracy reaches 23.2cm at the pedestrian walking speed.