アトラクタを用いた半自律掘削における人操作力のスペクトル解析に基づくオンライン軌道変形

抄録

We have developed a semi-autonomous remote excavation consisting of autonomy (by dynamical system with an attractor) and human action (by admittance control). There is a significant challenge that a trajectory of excavation is implemented to an attractor and it is deformed at the work site. For this purpose, the global stability of the attractor has to be guaranteed and it has to be appropriately deformed based on the operator’s intuition. In this paper, we design globally stable attractor based on teaching data and an intuitive online deformation method for the attractor ’s trajectory. Based on the spectral analysis of human inputs, the Weighted Least Square method is used. Moreover, for fast calculation, Recursive Least Square method is applied. The effectiveness of the proposed method is verified by 2-DOF simulations.