Automatic grasp planning algorithm for synergistic underactuated hands

Abstract

Over the last decades, effective grasp planning algorithms, which can handle grasps of common objects for robotic hands, have been extensively investigated. Among these, synergistic actuation, a promising approach, attracts lots of attentions. However, synergies will reduce the dexterity and manipulability of robotic hands. To obtain high grasp quality, the synergy parameters need to be optimized for different objects. This paper focuses on the automatic grasp planning algorithm for synergistic underactuated hands. Our major efforts are the quantitative description of the grasp quality and the optimization of synergy parameters for different objects. Based on the proposed virtual object model, the grasp quality function, equilibrium equations and contact constraints are defined. Then an automatic grasp planning algorithm is established for synergistic underactuated hands. Finally, numerical examples are presented on a robotic hand model, and simulation results show that the proposed automatic grasp planning algorithm can be used to grasp different objects.