《中国康复理论与实践》

《中国康复理论与实践》 ›› 2022, Vol. 28 ›› Issue (10): 1231-1240.doi: 10.3969/j.issn.1006-9771.2022.10.015

• 辅助技术 • 上一篇    

适应患者个体差异的上肢康复机器人直接示教技术

林高1,2,3,张道辉1,2(),赵新刚1,2   

  1. 1.中国科学院沈阳自动化研究所/机器人学国家重点实验室,辽宁沈阳市 110016
    2.中国科学院机器人与智能制造创新研究院,辽宁沈阳市 110169
    3.中国医科大学智能医学学院,辽宁沈阳市 110004
  • 收稿日期:2021-10-28 修回日期:2022-03-29 出版日期:2022-10-25 发布日期:2022-11-08
  • 通讯作者: 张道辉 E-mail:zhangdaohui@sia.cn
  • 作者简介:林高(1995-),男,汉族,广东湛江市人,硕士研究生,主要研究方向：末端牵引式上肢康复机器人人机交互技术。
  • 基金资助:
    国家自然科学基金项目(U20A20197);国家自然科学基金项目(U1813214);国家自然科学基金项目(61903360);国家自然科学基金项目(92048302);兴辽英才计划项目(XLYC1908030);辽宁省自然科学基金项目(2019-KF-01-06);中国博士后科学基金项目(2019M661155)

A direct teaching technology of upper limb rehabilitation robot meeting individual difference

LIN Gao1,2,3,ZHANG Daohui1,2(),ZHAO Xingang1,2   

  1. 1. State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
    2. Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang, Liaoning 110169, China
    3. School of Intelligent Medicine, China Medical University, Shenyang, Liaoning 110004, China
  • Received:2021-10-28 Revised:2022-03-29 Published:2022-10-25 Online:2022-11-08
  • Contact: ZHANG Daohui E-mail:zhangdaohui@sia.cn
  • Supported by:
    National Natural Science Foundation of China(U20A20197);National Natural Science Foundation of China(U1813214);National Natural Science Foundation of China(61903360);National Natural Science Foundation of China(92048302);Revitalizing Liaoning Talents Plan(XLYC1908030);Natural Science Foundation of Liaoning Province(2019-KF-01-06);China Postdoctoral Science Foundation(2019M661155)

RichHTML

PDF (PC)

摘要：

目的 针对不同患者上肢训练轨迹与训练强度存在差异性的问题,本文提出一种末端牵引式上肢康复机器人直接示教技术(示教与示教再现)。

方法 通过加入上肢重力补偿与约束步长转换控制,提出一种基于Moveit的直接示教,实现上肢康复机器人的稳定示教并用于训练。六维力传感器在采集到力/力矩信息后,在机器人操作系统(ROS)中经坐标变换、上肢重力补偿与约束步长转换算法,使患者上肢与机械臂末端能对康复治疗师拖动力的方向进行顺应性跟踪,同时记录示教轨迹。在ROS中利用Moveit编写示教再现节点,用于调节康复训练的训练次数、训练速度与休息时间,并进行训练。通过Moveit对示教轨迹信息的运动学求解与轨迹规划,康复机器人可精准地带动患肢进行往复训练,从而达到个性化训练的效果。

结果 示教者分别在水平面、矢状面、冠状面训练模式下拖动5例健康受试者的上肢完成相同的训练动作,在空间训练模式下完成任意的曲线运动。直接示教用时约4~7 s,得到的轨迹光滑且轨迹走向符合施力方向。因受试者的身高、上肢各部分长度等存在差异,5例受试者在完成相同的训练动作时存在轨迹和位置上的不同。示教者通过人机交互界面设置好合适的训练强度,对5例受试者进行各模式下的训练。受试者训练时动作规范且感觉舒适。选取最接近平均值的受试者3的轨迹给5例受试者进行训练,发现除受试者3以外,其他受试者均存在训练动作不规范、欠拉伸或过度拉伸等情况。

结论 末端牵引式上肢康复机器人直接示教系统能很好地帮助示教者简单、快速、稳定地示教出适合不同受试者的多自由度上肢训练轨迹,并通过再现轨迹进行不同强度的康复训练,具有操作简单、针对性强、训练精准的特点。

关键词: 末端牵引, 上肢康复, 快速示教, 人机交互, 个性化训练

Abstract:

Objective To develop a direct teaching technology (teaching and teaching reproduction) of an end-traction upper limb rehabilitation robot, in allusion to the difference between the upper limb training trajectory and training intensity of different patients.

Methods A direct teaching method based on Moveit was proposed, to realize the stable teaching of the upper limb rehabilitation robot, by adding upper limb gravity compensation and constraint step conversion control. After the six-dimensional force sensor collecting the force/torque information, the coordinate transformation, upper limb gravity compensation and constraint step conversion algorithm were used in Robot Operating System (ROS), so that the upper limb of the patient and the end of the mechanical arm could comply with the direction of the drag force of the rehabilitation therapist track, and recorded the teaching track at the same time. In ROS, Moveit was used to write a teaching and reproduction node, which was used to adjust the training times, training speed and rest time of rehabilitation training, and conduct training. Through Moveit's kinematics solution and trajectory planning for the taught trajectory information, the rehabilitation robot could accurately drive the affected limb for reciprocating training, thereby achieving the effect of personalized training.

Results The instructors dragged the upper limbs of five healthy subjects in the horizontal plane, sagittal plane and coronal plane training modes to complete the same training action, and completed any curve movement in the space training mode. The direct teaching took about four to seven seconds and the obtained trajectory was smooth and the direction conformed to the direction of force. Due to difference in the height of the subjects and the length of each part of the upper limbs, the five subjects had different trajectories and positions when completing the same training exercise. The instructor set appropriate training intensity parameters through the human-computer interaction interface and trained five subjects in each mode. The subjects had regular movements and felt comfortable during training. The trajectory of subject three, which was closest to the average value, was selected to train five subjects. It was found that all subjects except subject three had irregular training movements, unstretched or overstretched, etc.

Conclusion The direct teaching system of the end-traction type upper limb rehabilitation robot can well help the instructor teach the multi-degree-of-freedom upper limb training trajectory suitable for different subjects easily, quickly and stably. Rehabilitation training of different intensities is carried out by reproducing the trajectory, which has the characteristics of simple operation, strong pertinence and accurate training.

Key words: terminal traction, upper limb rehabilitation, fast teaching, human-computer interaction, personalized training

中图分类号: