Aiming at the problem that in pedestrian positioning and gait measurement based on multiple wearable inertial measurement units (IMUs),heading alignment usually depends on the utilization of magnetometers,expert intervention and joint hypothesis models,leading to complex operation process and non-guaranteed accuracy,the paper proposed a heading alignment method applicable to multiple IMUs mounted on lower limbs:a unified zero-velocity correction algorithm was designed for IMUs mounted on feet,shanks and thighs,and the law that pedestrians' feet touch the ground periodically was used when walking to construct zero-velocity observations for accurately restoring the motion trajectories of all IMUs;then,based on trajectory similarity,the heading alignment of multiple IMUs was realized;finally,the accuracy of trajectory and heading alignment were evaluated in outdoor open scene.Experimental result showed that alignment accuracy of heading would be positively correlated with trajectory accuracy of IMU in each position,and the average accuracy of heading alignment would be better than 1°(root mean square) when the walking trajectory length is more than 15m;in general,the proposed method would be simple and easy to implement,and could meet the requirements of human motion analysis and pedestrian positioning based on multiple IMUs.
关键词
穿戴式惯性传感器/航向对齐/行人航迹推算/人体运动分析/扩展卡尔曼滤波
Key words
wearable IMUs/heading alignment/pedestrian dead reckoning/human motion analysis/extended Kalman filter
维普
万方数据