Trajectory Generation for Adaptive Motion by Phase Feedback-Synchronization of Multicycle Human Movement

Abstract

Synchronous motion is one of the important ability for the co-operation work by human. The focus of our research is to develop a robust and adaptive synchronous trajectory generation method for the robot. To be able to follow the uncertain action by human co-worker, the trajectory generation method must be required adaptability to frequency and phase of movement of human co-worker. Key techniques for our method are PLL (Phase Locked Loop), Fourier series approximation, wavelet transformation for trajectory. PLL technique achieves the phase synchronization with an arbitrary cyclic motion, like as human walking. Fourier series approximation of target trajectory allows us description of wide variety cyclic motion for the robot. More over, we can select the synchronous frequency from human movement with multiple cycles by wavelet transformation. The experiment of synchronization with 3-DOF manipulator and human demonstrator are carried out. As a result, we confirm synchronous performance and the effectiveness of proposed method.