Fast Force Optimization of Multi-fingered Robotic Hand Grasps Based on Lagrange Multiplier Method

This paper presents a fast force optimization algorithm of multi-fmgered robotic hand grasps satisfying force closure equation and frictional constraints based on Lagrange multiplier method. It is based on the minimization of a cost function but the derived model does not ensure the satisfaction of the frictional constraints. A discrete iterative algorithm with realtime detection of contact safety threshold can modify the model until the internal force to cope with friction cone constraints. A very simple practical algorithm is developed which can be easily interpreted, and a very cheaply computable lower bound is given for the grasping force. It provides the suboptimal contact force to realize grasping by using Lagrange multiplier method. The proposed method is simulated on multi-fmgered hand grasps with hard and soft fmger types. The grasp examples indicate it can ensure the suboptimal grasping force to manipulate object. Simulation results demonstrate the efficiency of the nonlinear force optimization method.