121 筋電信号による動作識別とロボット制御(GS-2.15 人間・機械協調系)

概要

論文の詳細を見る
The number of amputees is increasing by industrial or traffic accidents, although safety control and accident prevention are made much of. Since amputees use prosthesis, the needs of prosthesis are increasing year after year. This work represents a research on multi-functionalization of myoelectric upper limb prosthesis. Pattern recognition motions of hands from Electromyogram (EMG) signals were attempted. The target function of each operation is given and EMG signals measured on the skin surface are separated the signal corresponding to each operation using linear multiple regression model. Using this technique, three motions (a grip, a knob, an opening of a hand) were recognized. The validity of the proposed signal separation by the proposed technique is verified by the experiment using three-fingered robot hand. In this experiment, EMG was measured, the model was created, and the robot hand was operated on-line. The result showed that high precision of recognition with this method.
社団法人日本機械学会の論文
2004-03-17