Hybrid Impedance Control of Robot Manipulators with Neural Networks Compensation

Abstract

This article presents an hybrid impedance control approach with neural network forthe compensation of robot manipulator's modelling errors. The proposed algorithm, consists of an outer hybrid impedance control loop that generates the reference acceleration to an inner inverse dynamics control loop. In order to improve the controller robustness, a compensation action of the manipulator modelling errors, is introduced, acting on the target acceleration. This compensation action is based on a neural network model achieved by minimising the modelling errors along the manipulator trajectory. The neural network algorithm uses an error training signal to model errors, that is minimised along the trajectory. The performance of the hybrid impedance control system with neural network compensation, is illustrated by computer simulations with a two degree-of-freedom PUMA 560 robot, which endeffector is forced to move along a frictionless surface located perpendicular to a horizontal plane. The results obtained, reveal an accurate force tracking and position control in robotic tasks, where is assumed significant uncertainties in the robot dynamic model.