Dynamic contact modeling of a soft fingertip and realization of dynamic object grasping for a multi-fingered robotic hand

我们扩展的动态稳定的物体抓取方法，由一对机器人的手指，我们已经提出了以前，由三个机器人的手指。通过这种扩展，可以完成一个重握任务，实现了一个激烈的目标姿态控制。提出了一种无需外部传感器的无传感器动态物体操纵方法，实现了物体的位置和姿态控制。提出了一种新的利用具有扭转关节的机器人手指系统进行动态物体抓取和操作的方法，并通过实验验证了该方法的有效性。提出了一种新的物体抓取模型，该模型中指尖轮廓和物体轮廓均为任意光滑曲线，并通过数值仿真验证了该模型的有效性。

We extend the dynamic stable object grasping method by a pair of robotic fingers, in which we have been proposed previously, to that by a triple robotic fingers. A regrasping task could be accomplished by this extension and it realized a drastic object attitude control. An external sensorless dynamic object manipulation method was newly proposed, which can realize an object position and attitude control without use of any external sensor. Another dynamic object grasping and manipulation method by using a novel robotic fingers system, in which each fingertip owns torsional joint, was newly proposed, and its usefulness was verified through experiments. A new object grasping model, in which both contour of fingertip and object are arbitrary smooth curve, was newly proposed, and its effectiveness was verified through numerical simulations.