Research On Key Technology Of Pedestrian Indoor Navigation System Based On MEMS/UWB

GNSS technology has achieved good navigation effect in outdoor environment,it can not perform the same performance in indoor where the signal is seriously occluded.In order to meet people's demand for indoor navigation,many indoor positioning technologies have emerged,such as INS,WIFI,Bluetooth positioning technology,etc.However,each positioning technology has its own defects,which cause people's needs for indoor positioning can not be satisfied.Based on this background,this paper studies UWB positioning technology and MEMS positioning technology,combines the advantages of the two technologies to achieve high-precision indoor navigation.The main research contents is as follow.Firstly,Aiming at the problems of false detection,missing detection and low fitness in traditional zero-speed detection methods,after analyzing the relationship between pedestrian gait characteristics and generalized likelihood ratio test,a zero-speed detection method based on generalized likelihood ratio curve geometric transformation is proposed.Firstly,the curve valley bottom is obtained by curve translation,and then a non-zero-speed interval is proposed by using the slope difference between the supporting phase and the swing phase in pedestrian gait.The algorithm can effectively reduce the rate of false detection,and has high adaptability to different pedestrian gait at different speeds.Based on this zero-speed detection method,the ZUPT algorithm is designed to improve the navigation accuracy of the system.Secondly,the propagation process of UWB signal is vulnerable to environmental interference,non-line-of-sight error and multi-path effect will lead to outliers in the positioning process,leading to discontinuous positioning.Based on the analysis of the traditional least squares method,this paper presents an algorithm of Unscented Kalman filter(UKF)for outliers rejection for maneuvering target tracking.The 3? statistical criteria is introduced into UKF,using a multi-dimensional bilateral truncation tail-type threshold vector function eliminates erroneous data in observation information and realizes modified UKF.The simulation and experimental results show that the algorithm can effectively eliminate outliers in the positioning process and improve the stability and reliability of the system.Thirdly,Analyzing the advantages and disadvantages of the positioning technology of the MEMS and UWB,and combining the advantages of these technologies,the problems of discontinuity,instability and error accumulation of the single positioning mode are solved.Using federated Kalman as the fusion tool,the UWB error data is further eliminated by using the combined positioning results.At the same time,the accumulative positioning errors of MEMS are corrected by using accurate UWB positioning information to further improve the positioning accuracy and reliability of the system.