Multi-fingered grasping force optimization based on generalized penalty-function concepts

Abstract

This paper presents an efficient multi-fingered grasping force optimization (GFO) method based on generalized penalty-function concepts. In view of the fact that the mainstream multi-fingered GFO method often treats the second-order cone programming (SOCP) problem as a semi-definite programming (SDP) problem, whose computational complexity is high, we hereby use the barrier function to construct the regularized optimization problem. The trade-off representation of different dimension objective functions is given, and the penalty factor is introduced to form the augmented optimization objective function. For specific operational tasks, by adjusting the penalty factor, a more compact, stable or slack, flexible grasping scheme could be obtained. Monte Carlo simulation is used to determine the probability of successful grasping when variability is introduced, and the robustness of the proposed method in the change of contact position and the friction coefficient between hand and object is verified. Experimental results and dynamic simulation are given, which show that the proposed algorithm outperforms the mainstream SDP method in execution time and iteration number, and the obtained force distribution has both continuity and distribution. Operational flexibility is instructive for practical applications.