本研究以分析慣性感測訊號來重建手臂運動軌跡，索量測訊號包含角速度與線加速度。研究分為兩部分，第一部份為技術發展，也就是肢體運動模擬與軌跡重建方法建構，運動模擬可利用運動學推導出角速度與線加速度方程式，以模擬肢體運動過程產生之相對座標角速度與線加速度值；而軌跡重建須用絕對座標之加速度，但要使相對座標數值轉換到絕對座標上則需靠轉移矩陣來轉換，轉移矩陣可由四元數法求得，而四元數值可由角位移得到，本研究提出一新穎方法獲得角位移，由角速度訊號之頻譜獲得頻域訊號之頻率、幅值及相位，從而計算其角位移；同理，絕對座標之線加速度，同樣經過頻譜分析獲得其頻域訊號之頻率、幅值及相位，以計算線位移，如此進而重建運動軌跡。第二部份為技術驗證，設計五種迴旋動作，並比較模擬訊號與實際量測之慣性感測訊號重建之軌跡正確性。 由模擬訊號與實驗數據之軌跡重建結果相比，所發展之軌跡重建計數據有相當之正確性。The aim of this study is to analyze inertia sensing signals including acceleration and angular velocity for the reconstruction of motion trajectory of upper limbs. This study consists of two parts. The first part is the technical develop, that is limbs motion simulation and build the method of trajectory reconstruction. In order to simulate the angular velocity and linear acceleration value on the local coordinates in limb movement, can derive the angular velocity equation and linear acceleration equation by using kinematics. However, reconstruction of trajectory will use the acceleration on global coordinates, so make the value on local coordinate can transform to the global coordinates by use transformation matrix. Then transformation matrix can get by using quaternion, and quaternion can get by using angular displacement. This study presents new method to obtain angular displacement, by calculated frequency domain signal frequency, amplitude and phase that from angular velocity signal spectrum. Similarly, by using the spectrum of linear acceleration on global coordinates, can compute the linear displacement, and reconstruct motion trajectory. Secondly, is the technical verification. Design five revolution motions, to compare the correctness of reconstruction of motion trajectory by used simulation signal and inertia sensing signal from actual measurement. Through comparing the results between simulated signals and experimental data, the developed motion trajectory reconstruction scheme is justified.
