引用本文
  •    [点击复制]
  •    [点击复制]
【打印本页】 【在线阅读全文】【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览 574次   下载 574 本文二维码信息
码上扫一扫!
腰绑式行人导航系统基于缓存区的启发式航向补偿算法研究
王鹏宇,孟之栋,邓志红
0
(北京理工大学自动化学院,北京 100081)
摘要:
腰绑式行人导航系统通过航位推算方法实现行人定位导航,对于实行灾难救援和单兵作战等任务十分便利。行人行进步长和航向的准确估计是实现航位推算的前提。根据腰绑式惯性传感器信号的特征,采用基于三轴合成加速度的峰值检测进行单步划分;利用步频和加速度方差信息构建线性步长模型。针对人员执行任务过程中由于运动方向不确定性导致传统启发式航向漂移消除算法过度修正甚至失效的问题,提出了一种基于缓存区的启发式航向补偿算法。根据运动方式选定N个复步作为一个缓存区,计算缓存区内相邻复步航向差值的方差,依据方差动态调整航向补偿算法的修正强度,避免过度修正。利用所提算法进行了矩形实验和操场实验,终点定位误差小于1%,表明了所提方法的有效性。
关键词:  行人导航  腰绑式  峰值检测  步长估计  航向补偿  缓存区
DOI:
基金项目:国家自然科学基金(61673067)
Research on Heuristic Heading Compensation Algorithm with Buffers in Waist-Mounted Pedestrian Navigation System
WANG Peng-yu,MENG Zhi-dong,DENG Zhi-hong
(School of Automation, Beijing Institute of Technology, Beijing 100081, China)
Abstract:
The waist-mounted pedestrian navigation system realizes pedestrian positioning and navigation by dead reckoning, which is very convenient for carrying out tasks such as disaster rescue and individual combat. Accurate estimation of pedestrian step length and heading is the premise of dead reckoning. According to the characteristics of the waist-mounted inertial sensor signal, the peak detection based on the tri-axial resultant acceleration is used for the single step division. The step frequency and the acceleration variance are used to construct the linear step length estimation model. In view of the problem that the traditional heuristic drift elimination method can lead to over correction or even fails due to the uncertainty of pedestrian's movement direction in the process of performing tasks, a heuristic heading compensation algorithm with buffers is proposed. According to the motion mode, N double-steps are allocated as a buffer to calculate the variance of the heading differences between the adjacent double-steps in the buffer area, which is used to control the intensity of heuristic heading compensation method to avoid over correction. The experiments of rectangular trajectory and playground trajectory are carried out using the proposed algorithm and the results show that the end point positioning error is less than 1%, which indicates the effectiveness of the proposed method.
Key words:  Pedestrian navigation  Waist-mounted  Peak detection  Step length estimation  Heading angle correction  Buffers

用微信扫一扫

用微信扫一扫