A Dual Neural Network for Bi-criteria Torque Optimization of Redundant Robot Manipulators

Abstract

A dual neural network is presented for the bi-criteria joint torque optimization of kinematically redundant manipulators, which balances between the total energy consumption and the torque distribution among the joints. Joint torque limits are also incorporated simultaneously into the proposed optimization scheme. The dual neural network has a simple structure with only one layer of neurons and is proven to be globally exponentially convergent to the optimal solution. The effectiveness of dual neural network for this problem is demonstrated by simulation with the PUMA560 manipulator.