中文摘要：

针对下肢负重外骨骼机器人与其穿戴者运动协调的问题,设计一种人体步态感知系统,对人体下肢关键部位的运动状态采集和预测。用6个MTI—30姿态传感器采集人体下肢的姿态数据;以ARM微处理器STM32F407为计算单元,对采集的步态数据解算、预测和传输;用非线性时间序列分析Takens算法预测人体下肢关键部位的旋转运动。实验结果表明：该系统功能稳定,能准确对人体下肢的步态数据采集和预测,预测结果稳定可靠,为外骨骼控制器提供可靠的参考信息。

英文摘要：

Aimed at problem of movement coordinating of exoskeleton with its wearer,a human gait sensing system is designed for aquistition and prediction of lower limbs＇ motion state of the wearer. The system uses six attitude sensors called MTI—30 to acquire human gait data. ARM microcontrollers known as STM32F407 are used as data computing units in the system. Acquired gait data are resolved,predicted and transmitted. Predict rotational motion of human legs by nonlinear time series analysis Takens algorithm. The result of experiments show that system has stable function and can accurately acquire and predict human gait data of human legs,the predicted result is stable and reliable,which provides reliable reference data for exoskeleton controller.