Abstract:

This thesis is aimed to improve robot hand grasping strategy, force control, and hand-arm system planning and control. It is divided into two parts: the first part generalizes the stable conditions for grasping and uses some assumptions to simplify the equations describing stable conditions to propose a new grasping strategy. We use a grasping strategy based on impedance control with fingertip tactile sensors to control NTU-Hand V. The second part of this thesis extends our examination of robot hand action to the full robot hand-arm system. We describe how to generate an ideal grasp trajectory and establish four stages in planning the hand-arm system motion. Next, we explain how to control the arm-hand system and our hardware system. Finally, we demonstrate the whole system and control loop by illustrating how our humanoid robot, NINO, with NTU-Hand V and NTU robot arm V, would open a door, combining grasping control with arm-hand planning. Applying this research to arm-hand development, we can implement robot hand-arm systems in mobile robots or humanoid robots to perform grasping or manipulating tasks to improve humans’ daily lives.