基於EMG訊號之上肢穿戴式外骨骼機器人控制系統開發

Abstract

本論文的目標是基於EMG訊號開發提供使用者生活輔助的外骨骼機器人控制系統。EMG訊號是肌肉收縮過程產生的生理訊號，含有動作意圖的資訊，本論文利用閾值分析法解讀EMG訊號中的使用者意圖，並且搭配電流控制器與位置控制器提出不同的控制策略，產生適當的輸出力矩移動上肢外骨骼機器人，完成對使用者日常生活的幫扶。實驗結果顯示所設計之控制系統能即時獲得使用者意圖，搭配不同控制器的控制策略，完成任務，展現此系統實際日常生活應用的可行性。

The main goal of this thesis is the development and experimentation of an EMG-Based control system of the upper-limb wearable exoskeleton. Electromyography (EMG) signal, as a physiological signal generated during muscle contraction, implicates the user’s intention. In this thesis, for moving assistance in daily lives, we propose using the position and current controllers to manipulate the exoskeleton based on user’s intention obtained by the threshold approach. Preliminary experimental results show that the proposed system can derive user’s intention in real time, which will then be used by proposed control strategies to generate corresponding torques to move the exoskeleton for assistance. It thus demonstrates the feasibility of the proposed system to be applied for daily activities.