基于人体上肢运动信息的助残手抓取意图识别方法

Recognition method of grasping intention of assistive hand based on motion information of upper limb

在助残领域,助残手如能很好地识别手残人的动作意图,则能帮助手残人恢复一定的运动能力。通过研究正常人上肢的动作规律,提出了一种基于人体上肢运动信息的助残手抓取意图识别方法。利用角速度传感器采集健康人系鞋带时的动作数据,分析上肢角速度的变化与上肢运动状态之间的联系。通过实验发现,在上肢的连续动作中,手指的抓取操作发生在手臂静止之后。因此,根据上肢的角速度数据,提取特征参数来表征上肢加速运动、减速运动、静止等运动状态,设计了助残手抓取意图的识别方法并应用于系鞋带实验中。实验结果表明,基于人体上肢运动信息的助残手抓取意图识别方法具有一定的可行性和适用性,识别准确率达到98%。

In the field of disability assistance,if the assistive hand can well recognize the movement intention of the hand-disabled person,it can help the disabled to recover a certain degree of athletic ability.A grasping intention recognition method based on the motion information of human upper limb is proposed by studying the motion rules of normal human upper limb.The angular velocity sensor is used to collect the movement data of healthy people when tying shoelaces,and the relationship between the change of the angular velocity and the motion state of the upper limb is analyzed.The experiments show that in the continuous movement of the upper limb,the grasping operation of the fingers occurs after the arm is still.Therefore,the characteristic parameters such as maximum value and difference value are extracted to represent the motion state of the upper limb such as acceleration,deceleration,and stillness.A method for identifying the intention of disabled hand to grasp is designed and applied to the experiment of tying shoelaces.The experimental results show that the method based on human upper limb motion information has certain feasibility and applicability,and the recognition accuracy is 98%.