基于MEMS惯性测量单元的多源信息自适应步数检测方法

发布时间：2019-01-28 11:55

【摘要】：针对基于MEMS惯性测量单元的行人航迹推算中步数检测方法仅利用单一的加速度信号检测精度较低的问题,提出一种多源信息自适应步数检测方法。该方法通过综合考虑人体运动过程中的加速度信号和角速度信号,根据不同的步态特征通过设定不同的自适应阈值条件实现步数的检测。虽然常规的峰值检测算法和固定阈值检测算法在单一步态下步数检测精度相对较高,但是对复杂运动状态下的步数检测精度很差,无法适用于真实的行人运动过程中步数的检测。然而多源信息自适应步数检测方法却能够在行人不同运动状态下精确检测步数,该方法明显优于常规的峰值检测方法和阈值检测方法。试验结果表明,本文提出的多源信息自适应阈值检测方法在行人不同运动状态下的步数检测精度可达98%以上。
[Abstract]:In order to solve the problem that only a single acceleration signal is used in pedestrian track estimation based on MEMS inertial measurement unit, a multi-source adaptive step detection method is proposed. By synthetically considering the acceleration signal and the angular velocity signal in the course of human motion, the method can detect the number of steps by setting different adaptive threshold conditions according to different gait characteristics. Although the conventional peak detection algorithm and the fixed threshold detection algorithm have relatively high precision in detecting the number of steps under a single gait, the accuracy of the detection of the number of steps in a complex motion state is very poor. It can not be used to detect the number of steps in the course of real pedestrian movement. However, the multi-source information adaptive step detection method can accurately detect the number of steps in different moving states of pedestrians, and this method is obviously superior to the conventional peak detection method and threshold detection method. The experimental results show that the multi-source information adaptive threshold detection method proposed in this paper can detect more than 98% of the steps of pedestrians in different motion states.
【作者单位】：北京航空航天大学仪器科学与光电工程学院;
【基金】：国家重大科学仪器设备开发专项基金项目(2012YQ160185) 国家高技术研究发展计划(2015AA124002) 国家自然科学基金(60421063)
【分类号】：TN96

【参考文献】