形状记忆合金驱动手指功能康复外骨骼设计

为了研制适应关节黏弹特性、轻便、可穿戴的手指功能康复装置，在分析手指关节、肌腱运动机理的基础上，设计形状记忆合金（SMA）丝驱动的手指功能康复装置.建立该装置的运动学模型和SMA驱动模型，提出基于SMA丝电阻反馈的模糊神经网络PID控制方法，研制手指功能康复外骨骼样机，进行样机的运动性能和控制性能实验.结果表明，仿生外骨骼实现了预期的手指被动康复运动，拇指、食指和中指的最大弯曲角度与人体手指的弯曲角度相近，分别为130.5°、236.4°、242.5°；仿生外骨骼能够辅助手指完成日常抓握动作；模糊神经网络PID相比于传统PID控制，有效缩短了仿生外骨骼的响应时间，手指的屈伸康复频率为6次/min.

Design of exoskeleton for functional rehabilitation of fingers driven by shape memory alloy

A finger rehabilitation device driven by shape memory alloy (SMA) wire was designed based on the analysis of the movement mechanism of the finger joints and tendons, in order to design a light and wearable finger rehabilitation device that adapts to the viscoelastic properties of finger joints. Besides, a kinematics model and an SMA drive model of the device were established. A fuzzy neural network PID control method was developed based on resistance feedback of SMA wires. The prototype of the device was made, and experiments were conducted to test its motion function and control performance. Results showed that the bionic exoskeleton could drive the fingers to achieve the expected passive rehabilitation. The maximum bending angles of the thumb, index finger and middle finger were 130.5°, 236.4°, and 242.5° respectively, which were similar to the bending angles of human fingers. And the bionic exoskeleton was able to assist the fingers to complete daily grasping movements. Compared with the traditional PID control, the fuzzy neural network PID control could shorten the response time of the bionic exoskeleton substantially, and the frequency of finger flexion and extension movement up to 6 times per minute.