Research And Design Of Indoor Positioning System Based On Bluetooth

With the official release of Bluetooth Core Specification 5.1 in 2019,the direction finding function is added to the standard,which can help the device to acquire the direction of the Bluetooth signal and hopefully realize accuracy at the sub-meter or even centimeter level of the Bluetooth positioning system.Based on the CC2640R2 F chip and antenna array of Texas Instruments,the design and research of Bluetooth positioning system are carried out in this paper.Firstly,a distance measurement method based on TOF is presented,through oversampling,we can improve the accuracy of TOF.The measurement method of AOA angle is based on the principle of phase interferometer,and several reasons affecting the accuracy of AOA are analyzed in detail through differential equation,including the phase measurement accuracy,the direction of incoming wave,and the ratio of antenna array element spacing to carrier wavelength.The method of establishing error distribution function is proposed.For complex location algorithm,the corresponding error distribution map drawing method is proposed.The error distribution function can be fitted according to the error map.Secondly,in the data processing and fusion part of the attitude and heading reference system,the rotation matrices expressed by quaternion and Euler angle are established,and the attitude measurement formulas of accelerometer and magnetometer are derived.According to the relationship and the corresponding relationship between the two rotation matrix expressions,the quaternion initialization formulas are given.The attitude estimation algorithm based on traditional Mahony complementary filter and second-order adaptive EKF is designed.The simulation and experiment verify and compare the effectiveness of the algorithms.Then,in the part of Bluetooth location algorithm,I design weighted least squares and EKF location estimation algorithm which only depend on location information.In order to take smoothness and lag into account as much as possible,the integrated navigation model based on Bluetooth/AHRS is established,and the corresponding simulation experiments are carried out.The simulation results show that the integrated navigation can effectively improve the positioning accuracy and reduce the delay caused by conventional tracking filtering.At the same time,the geomagnetic coordinate system and navigation coordinate system can be estimated extra.The calibration of angle between coordinate systems arranged by the new system can be omitted.Finally,an experimental platform is built and a data visualization software based on general protocol framework is developed.The actual positioning system based on AOA triangular positioning mode of two anchors is realized.With UWB positioning data as a reference,the positioning accuracy of Bluetooth is calculated to be 0.613 m,which is greatly improved compared with traditional Bluetooth positioning accuracy.The experimental results of integrated navigation show that the effect of Bluetooth/AHRS integrated navigation depends on the accuracy of acceleration,while the low-cost MEMS is difficult to meet the requirements due to its low sensitivity and severe drift,especially in the case of free attitude.