A force control approach of robotic manipulators in non-rigid environments

In this paper, the application of an impedance control scheme to drive a robot with a desired force profile in the presence of non-rigid environments is considered. Fine force control obtained with the usual impedance based control schemes reveal some difficulties especially for non-rigid environments. Moreover, robot dynamic model and environmental uncertainties, can seriously affect the force tracking performance. In this paper, alternative control methodologies like predictive control and neural network compensation are used to design an impedance force control with an accurate force tracking performance. Thus, a predictive algorithm is designed to compute the virtual trajectory for the impedance controller. Further, the constrained acceleration produced by the impedance controller is compensated by an on-line neural network scheme in order to reduce the force errors. The performance of the proposed force controller is illustrated by simulation for a two degree-of-freedom PUMA 560 robot, which end-effector is forced to move along a surface on the vertical plane. Simulation results reveal an accurate force tracking performance for the proposed control scheme in the presence of non-rigid environments.