肌電訊號(electromyography)是人體的肌肉在收縮過程中所產生的生理訊號，此訊號會依人的動作不同而有不同的輸出電壓，因此可以藉由此訊號控制機械外骨骼或是估測人的意圖。 本研究透過兩通道的表面貼面電極，搭配自製的肌電訊號截取電路，量測大腿的股直肌與股外側肌兩塊肌肉的肌電訊號，再搭配均方根、平均絕對值及波形長度，對每塊肌肉萃取出三項特徵，再通過極限梯度提升法(eXtreme Gradient Boosting, XGBoost)建立與角度之間的模型;Electromyography is a physiological signal generated by the muscles of the human body during the contraction process. This signal will have different output voltages depending on the person′s actions.Therefore, the signal can be used to control the mechanical exoskeleton or estimate the person′s intention. In this study , the rectus femoris and lateral femoris muscles electromyographic signal were be measured by two-channel surfacemounted electrodes with a self-made ectromyographic signal interception circuit and extracted the feature signals with the root mean square, mean absolute value and waveform length . Then use these feature signals to establish a model with eXtreme Gradient Boosting in order to predict angle.
