Research On Key Technologies Of Indoor Positioning System Based On UWB

Ultra-wideband technology has the superiorities of high security,rapid transfer speed and powerful anti-interference capability,and has been widely used in indoor positioning system.However,the clock asynchronization in the terminal of UWB system will lead to positioning errors.In addition,when the target is in nonlinear motion,the commonly used target tracking algorithms have the matters of higher computational complexity and lower tracking accuracy.Therefore,this paper studies the key technologies of UWB indoor positioning system,and the specific work is as follows:(1)The theoretical basis of UWB indoor positioning is introduced.This paper describes the commonly used indoor positioning methods,target tracking algorithms and positioning performance indicators,and compares the positioning performance of Fang algorithm,Chan algorithm and Gaussian Newton algorithm in Matlab.(2)Aiming at the problem that anchor and moving tag clock are out of synchronization in TOA based positioning method,a clock calibration method based on differential positioning technology is presented.On the basis of TOA positioning method,a calibration label is added,and the clock difference between the unpositioned label and the calibration label is obtained through the calibration test,and then the clock difference is compensated to the TOA positioning and ranging to reduce the positioning error.In Matlab,the positioning results of rectangle,fork and oval motion track labels are simulated under the condition of fixed clock difference,linear clock difference,nonlinear clock difference and so on.It is demonstrated that the presented clock calibration method is able to compensate the errors caused by clock asynchronization and improve the positioning accuracy.(3)A robust extended Kalman filter algorithm β-EKF is proposed to solve the shortcomings of nonlinear target tracking algorithms.Based on the output of Taylor’s approximate nonlinear function,the error between the estimated and actual observed values is controlled by the robust threshold value to realize the adaptive correction of the target state estimation,improve the robustness of the model to noise and improve the accuracy of the state correction.In Matlab,the tracking simulation of three kinds of nonlinear moving trajectory targets is carried out.The results show that the proposed algorithm can obtain positioning results with higher precision and more robust than the extended Kalman filter,adaptive fading biased extended Kalman filter,untraced Kalman filter and other algorithms.Moreover,compared with the optimal untraced Kalman filter algorithm,The relative error is reduced by 29.1%,31.46% and 18.85%.(4)To address the matter that anchor has higher demand on clock synchronization in TDOA localization method,a passive tag-oriented multipath positioning method is proposed.Firstly,the multipath propagation mechanism between the anchor node and the tag signal is established.Then,the multipath flight time difference of the single transmitting source signal is used to calculate the tag position.Finally,the proposed β-EKF algorithm is used to filter the positioning results.Through theoretical analysis and simulation verification,the proposed method not only equates with the TDOA method in the Cramer-Rao Lower Bound,but also outperforms the Asynchronous Parametric Ranging,Double Side Two-Way Ranging and TDOA method in terms of communication efficiency.