基于MEMS惯性传感器的高精度步长估计算法
路永乐,陈永炜,邸 克,郭俊启,余 跃,刘 宇
【摘 要】摘要:针对传统计步算法精度低适用性不高的问题,提出了一种基于MEMS足部定位模块的高精度自适应步长估计算法。首先通过分析行人步态的周期特征,精确地判断每个步态周期中单脚离地和落地的时间差值,采用零速更新法结合时间阈值、加速度阈值和加速度方差阈值进行步态检测;然后对转换后地理坐标系下传感器信号进行双重积分,计算出行人每一步的空间轨迹,无需进行前期步长数据采集或者生物模型参数提取,即可实现对行人步长的精确估计。在AHRS_box惯性传感器模块上进行了实验,结果表明该算法估计得出的步长和行走距离误差均小于2.5%,证实了该算法的有效性。 【期刊名称】中国惯性技术学报 【年(卷),期】2024(026)002 【总页数】6
【关键词】行人导航系统;步长;步态;零速更新;MEMS惯性传感器 In the future, the measurement and monitor of temporal and spatial information of the pedestrian gait is essential for self-contained pedestrian outdoor and indoor navigation system and emergency rescue[1-2].Global Navigation Satellite System, such as Global Position System (GPS) and Globalnaya Navigatsionnay Sputnikovaya Sistema (GLONASS), has been playing an important role in pedestrian outdoor navigation system.The satellite signals are unable to penetrate buildings so that the indoor navigation techniques, such as ZigBee,Wi-Fi, Beacon,