低频压电马达的工作机理与动力学研究

通过检测人体行走步数，结合步距和行走方位，可推算出具体行走位置，从而在无GPS情况下，达到提高个人惯导位置精度的目的。为提高计步精度，利用加速度信号，首先采用5点平滑滤波，其次搜索自适应时间窗内的极值来求取自适应双阈值，最后检测信号向上穿越中阈值和向下穿越低阈值时刻。并结合动态精度判段停止或行走，从而缩短算法检测时间。实验结果表明，基于自适应时间窗、自适应双阈值和动态精度的步数检测算法能有效提高检测精度。

Study on Kinetics and Mechanism of Low Frequency Piezoelectric Motor

The specific position of walking can be deduced and the personal INS position accuracy without GPS can be improved by detecting human walking step numbers and combining with step distance and walking direction. In order to improve pedometer accuracy, the 5 point smoothing filtering is adopted by using acceleration signal firstly, and then the adaptive dual threshold are calculated with the extremes searched in the adaptive time window, finally, the time signals of the up through the middle threshold and down through the low threshold are detected, and the stop or walking is determined by the dynamic precision, thus the time of detection algorithm is shortened. The experimental results show that the pedometer detection algorithm based on the adaptive time window, adaptive dual threshold and dynamic precision can effectively improve the detection accuracy.