Abstract
It is important for robots to obtain appropriate grasp positions autonomously to make stable grasps. This paper proposes an automatic generation algorithm for getting the optimum planar grasp. Each finger is replaced with a two-dimensional linear spring model. Grasp stability is evaluated by using the eigenvalues of the grasp stiffness matrix derived from the potential energy of the grasp. To decrease the number of searching iterations, we derive the partial differentiation of the matrix by the contact position parameters. Effectiveness of our method is demonstrated through numerical examples.
