本論文提出以肌電訊號步態分類器之人形機器人步行控制，論文架構可分為肌電訊號擷取、步態周期分類及步態辨識三個部分。肌電訊號擷取完整步態周期之肌電訊號。在步態周期分類部分以肌電訊號最大值發生處區分出腳跟離地期、搖擺初期及腳跟接觸期。步態辨識使用下肢肌電訊號的功率頻譜密度及步態周期發生之時間做為特徵值並使用類神經網路鑑別步態周期的腳跟離地期、搖擺初期及腳跟接觸期，訓練及測試的結果，類神經網路可正確辨識上述動作。最後將類神經網路所分析出來的步態動作做為控制雙足機器人動作，驗證所提出的以步態周期時間參數來輔助下肢肌電訊號辨識是為可行的。

This thesis proposes an electromyography (EMG) based gait segmentation system for controlling the humanoid robot locomotion. The architecture of this thesis is classified into three stages that include signal acquisition, gait segmentation and gait recognition. In the signal acquisition stage, the EMG signal was collected for forming a complete gait cycle. In the gait segmentation stage, the signal in a complete gait cycle is further divided into standing phase and swing phase. In the standing phase, the duration of heel contact and heel off standing is segmented by evaluating the maximum signal; in the swing phase, early swing period is segmented by the same method. In the gait recognition stage, power spectrum density, gait cycle time is used as the feature for gait recognition. In this thesis, artificial neural network (ANN) based gait cycle recognition is developed for recognizing the gait pattern according to the input EMG features. Finally, the experiments results were discussed to demonstrate the capability of the proposed approach.

[1] M. A. Oskoei and H. Hu, “Myoelectric control systems - A survey,” Biomedical Signal Processing and Control, vol. 2, no. 4, pp. 275 – 294, 2007.
[2] B. Hudgins, P. Parker, and R. N. Scott, “A new strategy for multifunction myoelectric control,” IEEE Transaction on Biomedical Engineering, vol. 40, no. 1, pp. 82 – 94, 1993.
[3] J. Zhao, “EMG control for a five-fingered under actuated prosthetic hand based on wavelet transform and sample entropy,” IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 3215 – 3220, 2006.
[4] R. Kinnaird and P. Ferris, “Medial gastrocnemius myoelectric control of a robotic ankle exoskeleton,” IEEE Transaction on Neural System and Rehabilitation Engineering, vol. 17, no. 1, pp. 31 – 37, 2009.
[5] H. Huang, A. Kuiken, and D. Lipschutz “A strategy for identifying locomotion mode using surface electromyography,” IEEE Transactions on Biomedical Engineer, vol. 56. no.1, pp. 65 – 73, 2009
[6] S. Lee and N. Saridis, “The control of a prosthetic arm by EMG pattern recognition,” IEEE Transactions on Automatic Control, vol. 29, no. 4, pp. 290 – 302, 1984
[7] 鍾睿洲，『具重心高度調適之雙足線性倒單擺步行控制』，碩士論文，臺灣科技大學，民國100年。
[8] 王耿良，『具適應斜坡行動能力之雙足步行機器人設計開發』，碩士論文，長庚大學，民國92年。
[9] 黃千芳，『以仿生學角度探討類人機器手指關節運動行為分析』，碩士論文，長庚大學，民國96年。
[10] 羅華強，“類神經網路－MATLAB 的應用”，高立圖書，民國100年。
[11] 郭怡良，“基礎肌動學”，台灣愛思唯爾，民國102年
[12] 賴昆城，“人體解剖生理學”，華騰文化, 民國101年。
[13] Christian Fleischer and G‥unter Hommel, “A human-exoskeleton interface utilizing electromyography,” IEEE Transactions on Robotics, vol. 24, no. 4, pp. 872 – 882, 2008.
[14] K. Kiguchi and Y. Imada, “EMG-based control for lower-limb power-assist exoskeletons,” IEEE Workshop on Robotic Intelligence in Informationally Structured Space, pp. 19 – 24, 2009.