Research On Indoor Positioning Technology Based On UWB/PDR Combination

In today's era of rapid development of the Internet,a lot of changes have taken place in the way of human life,human demand for location applications is growing.At present,the development of global satellite navigation system has been relatively mature,which basically meets the needs of human outdoor location.However,in the indoor and tunnel sheltered environment,GNSS satellite signal is covered by obstacles,so it is difficult to achieve positioning.Therefore,in order to meet the needs of human location in the indoor environment,relevant scholars have carried out a series of research on electromagnetic wave,light,sound and other sensors.In recent years,relevant researchers have proposed indoor positioning technology based on Bluetooth,Ultra Wideband,Wi Fi and so on.Each sensor has its own advantages in the positioning process.However,due to the complex and changeable indoor environment,only one sensor can not meet the needs of human indoor positioning.In this context,this paper systematically studies the UWB positioning technology and the pedestrian dead reckoning technology.Combined with the positioning characteristics of the two sensors,the combination of the two sensors is proposed.The combined positioning model is studied in detail and verified by experiments.The main contents include:(1)In the process of UWB positioning,the signal is easy to be interfered by the indoor complex environment,which leads to the degradation of ranging accuracy.This paper collects several groups of training data,establishes the UWB ranging error correction model based on exponential function,and carries out the experimental verification.The experimental results show that the average ranging accuracy is improved by 60% after the error model correction,which shows that the error correction model can effectively reduce the UWB ranging error.In the static positioning experiment,the positioning accuracy of the UWB positioning algorithm based on gradient descent method is 50.61% higher than that of the least square positioning algorithm;In the dynamic positioning experiment,the positioning accuracy of UWB algorithm based on gradient descent method is 17.21% higher than that of least square algorithm.(2)In the process of pedestrian dead reckoning and positioning,the results will have accumulated errors over time.In this paper,combined with the characteristics of pedestrian movement,a gait detection model based on multi parameter constraints is established.The acceleration is constrained by the parameters of acceleration peak,trough and time difference,Compared with the peak detection model,the accuracy of the proposed gait detection model is improved by 83.98% in walking state and 86.12% in running state.Because the pedestrian heading angle is easy to deviate from the real heading in the middle and later stage of positioning,this paper proposes a method to constrain the heading angle by using indoor map information.Experimental results show that the average error of heading estimation with map constraint is 74.44% lower than that of the original heading angle error.This method improves the accuracy of heading estimation in the process of pedestrian movement,and provides a guarantee for the subsequent combined positioning.(3)The positioning accuracy of EKF combined positioning model will gradually converge with the passage of time,resulting in the influence of the data of the previous moment on the state estimation of the next moment will be reduced.This paper studies and analyzes the characteristics of two kinds of sensor positioning,establishes the indoor UWB /PDR combination adaptive EKF positioning model,uses the fading factor obtained by fading filter to adaptively adjust the estimation value of prior probability,and combines with EKF to realize positioning.The experimental results show that the positioning accuracy of the adaptive EKF combination model is 50.11% higher than that of UWB,83.32% higher than that of PDR,and 28.98% higher than that of EKF combination model.