【摘要】 通过MEMS惯性传感器采集上肢运动的姿态角数据，分析姿态角数据表现出来的运动规律，提出了一种基于意图识别的助残机械手控制方法。该方法包括两部分，第一部分计算上肢运动过程中姿态角的移动方差综合值来判断手臂的动静状态；第二部分首先记录预备阶段后手臂第一次静止状态时的姿态角数据，然后将之后手臂每一次静止状态的姿态角数据与第一次静止状态的姿态角数据进行比对，来判断机械手的开闭意图。将此方法应用于日常生活中常见的五种抓取动作，识别准确率达到94%,表明该方法具有一定的可行性和普适性。更多还原

【Abstract】 The attitude angle data of upper limb movement are collected by MEMS inertial sensor, the motion law shown by the attitude angle data is analyzed, and a control method of disability manipulator based on intention recognition is proposed. The method includes two parts. In the first part, the dynamic and static state of the arm is estimated through the calculation of the moving variance of the attitude angle in the process of upper limb movement. In the second part, the attitude angle data of the arm in the first static state after the preparation stage are first recorded, and the attitude angle data of each following static state of the arm are then compared with those of the first static state to judge the opening and closing intention of the manipulator. This method is applied to five common grasping movements in daily life, and the recognition accuracy is 94%,which shows that this method has certain feasibility and universality.更多还原