Reference compensation technique of neural force tracking impedance control for robot manipulators

Abstract

In this paper, the neural impedance controller is formulated to regulate the contact force with the environment. When robot uncertainties are present, the performance of the impedance controller is degraded. To compensate for uncertainties in both robot dynamics and environment, neural network is introduced at the desired trajectory. The training signal is defined to satisfy the desired goal. This leads to the remarkable advantage of no requirement of modifying an internal force control structure. Neural network actually compensates for uncertainties at the input trajectory level in on-line fashion. The robust position and force tracking performance of a robot manipulator is confirmed by simulation studies.