G101013 EMG Control of Power Assist Glove using Self-Organizing Maps

Abstract

In this study, a finger and wrist motions are discriminated using EMG on forearm to move the power assist glove, which can assist grasp. A finger and a wrist are moved when various muscles on a forearm contract simultaneously. Therefore, a finger and a wrist motions must be discriminated to detect a human intention to move a finger. By introducing Self-organizing maps (SOM), discrimination of these motions can be improved. In this paper, the effectiveness of discrimination method using SOM is verified experimentally.