基于INS/UWB紧组合的行人室内定位方法

打开文本图片集

中图分类号：TH89 文献标志码：A DOI：10.12305/j.issn.1001-506X.2025.04.26

Abstract：In view of the problem of poor pedestrian positioning accuracy in the indoor complex application scene，the inertial navigation system/ultra-wide band（INS/UWB） tight integrated positioning algorithm is proposed. Taking one gait cycle of foot motion as a basic search unit and by analyzing pedestrian's foot motion， z axis angular velocity information is used to detect zero-speed state under composite motions of pedestrians accurately，so as to further suppress inertial solution accumulative error with zero velocity update（ZUPT）. Meanwhile，considering the impact of non-line-of-sight（NLOS）on UWB ranging accuracy，an NLOS error model is established using a backpropagation neural network adaptive learning method to compensate for UWB ranging erors. The experimental results show that the INS/UWB tight integration algorithm proposed in indoor complex application scene has a positioning error of 0.19m ，which is 84.6% higher than the INS/UWB tight integration positioning method without ZUPT and NLOS error compensation.

Keywords：pedestrian indoor positioning；ultra-wide band（UWB）；tight integration；zero velocity update （ZUPT）；non-line-of-sight（NLOS） error

0引言

随着物联网技术的发展，基于位置服务定位（locationbasedservice，LBS）的导航定位系统在人们生活中扮演着越来越重要的角色[1-5]。（剩余18227字）