全上肢外骨骼康复机器人与人机协同智能训练方法研究

针对中风导致的上肢运动功能障碍,线驱动外骨骼机器人在康复训练中具有柔顺性与安全性高等优势。但对于健侧颈7神经移位(CSCNT)术后康复需求,其在机构设计、柔顺控制、康复训练与康复评价方面仍有较大改进空间,需进一步深入研究。本项目拟开展可互换全上肢外骨骼机构设计,实现包括腕、手的全上肢康复训练,并可完成外骨骼在健侧和患侧的便捷互换；基于sEMG构建人机共融刚度模型,设计变刚度柔顺控制策略,解决柔顺性和刚度之间的平衡问题；提出"模仿-同步-合作"人机协同康复模式,开展基于视觉演示学习的智能训练方法研究,实现机器人化康复训练;研究面向人机协同康复模式的多维度运动能力评价方案,对各康复模式差异化精量评价；最后,搭建实验平台开展临床试验研究。项目成果不仅为CSCNT术后患者研制全上肢外骨骼机器人,而且对于类似外骨骼机器人的人机共融控制与机器人化智能训练等共性问题的研究,具有重要的科学意义和学术价值。

In rehabilitation training for the motor dysfunction of upper limbs caused by stroke, cable-driven exoskeleton robots have the advantages of high compliance and safety. However, for the rehabilitation needs after the Contralateral Seventh Cervical Nerve Transfer（CSCNT）, there is still possibility of improvement in the mechanism design, compliant control, rehabilitation training and rehabilitation evaluation, which needs further study. In this project, we intend to carry out the design of an interchangeable exoskeleton mechanism for the whole upper limb, so as to realize the rehabilitation training of the whole upper limb including the wrist and the hand, and to realize convenient interchange of the exoskeleton on the paralyzed side and the nonparalyzed side. Using sEMG, the human-machine fusion stiffness model and the variable stiffness compliant control strategy will be constructed to balance compliance and stiffness. We plan to put forward an "imitation-synchronization-cooperation" human-machine collaborative rehabilitation mode and carry out research on the intelligent training method based on visual demonstration learning to realize the robotic rehabilitation training. We will study the multi-dimensional motion ability evaluation scheme oriented to human-machine collaborative rehabilitation modes so that different rehabilitation modes can be evaluated diversely and precisely. Finally, the overall experimental platform will be built to carry out clinical rehabilitation experiments. This project will not only develop the whole upper limb exoskeleton robot system for rehabilitation training after CSCNT, but also have important scientific significance and academic value for the research on the common problems such as the human-machine fusion control of similar exoskeleton robots and intelligent robotic rehabilitation.