多手指人工義手系統研發

本篇論文主要建構了一套以形狀記憶合金做為驅動器之電子手，並結合肌電訊號辨別系統，進而發展出一套以肌電訊號控制的人工義手系統。整體系統架構可包括形狀記憶合金電子手本體、以及由個人電腦與NI-CompactRIO 所組成的肌電訊號辨別系統兩大部分，第一部份介紹人工義手的架構，包含各關節之可動性與形狀記憶合金在義手中的驅動方式；第二部份則是簡單介紹人體手臂前肢的肌肉構造與電極貼片的黏貼處以及訊號擷取的處理過程，其中包含訊號的放大、濾波、整流等步驟，達到取得乾淨的肌電訊號的目的，過程以三組貼於手臂之電極貼片擷取肌電訊號，每組肌電訊號擷取六個特徵參數，並採用支持向量機做手勢之辨識。本篇研究不但證實形狀記憶合金其取代馬達的可能性，並驗證了即時手部動作辨識系統之可行性與有效性。

關鍵字

形狀記憶合金；肌電訊號；支持向量機

並列摘要

In this thesis we use shape memory alloys to drive a electronic hand, combined with the EMG signal to identify the command, and then develop an EMG controlled prosthetic hand system. There are two parts in the overall system architecture, including the shape memory alloy hand, and EMG signals to identify the commands using a PC and the NI-CompactRIO system. The first part describes the architecture of the artificial prosthetic hand, including the joint mobility and the way how shape memory alloys drive the fingers; the second part considers the forelimb muscle structure of a human arm with surface electrodes attached for EMG signal acquisitions, which is then processed by signal amplification, filtering, rectification, and other steps. To extract features of EMG signals, six characteristic parameters are used, and support vector machine is employed for hand gesture classification. This study not only confirmed the possibility of using shape memory alloys to replace the motor, but also verified the feasibility and effectiveness of the developed EMG driven SMA prosthetic hand.

並列關鍵字

SMA(Shape Memory Alloy) ； Electromyographic ； Support Vector

參考文獻

[2] Takashi Mineta, Toshiaki Mitsui, Yoshiyuki Watanabe, Seiya Kobayashi and

Youichi Haga "An active guide wire with shape memory alloy bending actuator

pp.632-637, 2002.

[3] J. Abadiea, N. Chailleta and C. Lexcellent, "An integrated shape memory alloy

micro-actuator controlled by thermoelectric effect", Sensors and Actuators, A99,

國際替代計量

多手指人工義手系統研發

全文下載

主題瀏覽