基於肌電波之人手與義肢控制

Abstract

隨著醫學、生物科技、電機資訊等領域的快速發展，我們對人體也因此有了更深層的了解，不禁讓研究人員思考以腦波、肌電波等生理訊號為基礎，建立新的人機介面的可能性，加上近年來由於職業災害與意外等等因素，導致肢體殘障人數數量的大幅增加，義肢的需求也日益增加，因此發展一有別於傳統的神經與肌肉系統之通訊與控制管道，讓神經或肢體受創的人有另一種與外界溝通的新選擇，勢必是一新的趨勢，因此我們希望藉由對肌電義肢的研究，能對肌電訊號在機器臂的控制上有更進一步的認識。基於此構想，我們計畫建立類似的義肢系統作為我們的工作平台，以實現視覺回饋的義肢控制作為初步的目標，希望能找到簡單的方法得到穩定的肌電波，用以控制義肢，進而實現建立新人機介面的目標，並將此人機介面應用到機器臂的控制上，使其成為人手的延伸。因此我們選擇具有設計簡單、重量輕及價格合理等優點的德國OTTO BOCK義肢以及國內的德林義肢作為實現的對象，而我們所實現的基於肌電波之人手與義肢控制系統經過實驗的印證，更確定了肌電波在機器臂控制應用上的可能性。

Along with fast development of biomedical and electrical engineering, we know our body functions much better than before. It is now possible to build a human machine interface based on electroencephalogram and electromyography. Meanwhile, the demand of prosthesis increases due to more people hurt during work or accident. In this thesis, we intend to study the prosthesis that is controlled by electroencephalogram or electromyography. We will first propose a human hand prosthesis control system based on electromyography, which is similar to the prosthesis control system. We will develop a new human machine interface and apply it for robot control. We experiment with five males who are 20 to 30 years old. We will prove system accuracy with human adaptation. Everyone exercise for 5、10 and 15 minutes respectively before experiment. The result is 92%、96% and 98% accuracy for 5、10 and 15 minutes exercise respectively. It is proving that the accuracy depends on human adaptation and this system is high performance with enough exercise time.