Developing a Novel Bilateral Arm Training on Rehabilitation Robot NTUH-II with IMU Measurement System for Neurological and Orthopedic Disorders

• Main Authors: Wei-Ming Lien, 連惟忞
• Other Authors: Li-Chen Fu
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/85192825269480616727

碩士 === 國立臺灣大學 === 電機工程學研究所 === 103 === Patients with upper limb disabilities caused by neurologic or orthopedic disorder required duly rehabilitation to regain their motor function and prevent complications. To reduce the burdens on both physical therapists and patients, in this research, based on a novel physical therapy called bilateral arm training (BAT), which is one of self-rehabilitation capable of promoting recovery of the impaired side with the help from the intact side, we develop a bilateral control interface for our recently developed 8 degrees-of-freedom (DOFs) exoskeleton type upper extremity rehabilitation robot, named NTUH-II. To extract motion information of the intact arm, we employ our self-developed inertia motion unit (IMU) Measurement System. It turns out that the unique and compact feature of NTUH-II together with IMU, which is its especial large range of motion (ROM) for shoulder, can allow patients to and practice the activities of daily living (ADL) BAT extremely effectively. More specifically, the control strategies are to enable the patient’s impaired or injured limb to undertake passive or active-assistive rehabilitation therapy through Robot-assisted Bilateral Arm Training (RBAT) while letting IMU measure the voluntary motion of the healthy arm. As for the active-assisting strategy, it is to generate the guidance torque as a sufficient aid to the impaired side if needed with reference to the motion of the healthy side. In addition to the self-motion control, leader-follower control is also available on proposed control scheme, where the therapist is allowed to easily generate the desired motion profile for the motion on patient’s impaired side, and can satisfy various requirement in robot-assisted rehabilitation. Compared to the conventional approaches which are limited to specific types of pre-defined reference profile, the integration of the proposed framework enables us to generate and realize more sophisticated motion trajectories. Various experiments have been conducted on eight healthy subjects to verify the tracking performance and assisting effect of the proposed bilateral control, and the results show that proposed control scheme can boost the subject physical coordination with less effort. From these results, the utility of the developed bilateral control method has been confirmed.