Neural network parallel force/position control of robot manipulators under environment and robot dynamics uncertainties

The performance of a parallel force/position controller for robot force tracking is affected by the uncertainties in both the robot dynamics and the environment stiffness. This paper aims to improve the controller's robustness by applying the neural network (NN) technique to compensate for the robot dynamics at the input trajectory level and adaptive feed-forward compensation to cope with variations in the contact environment. A NN control technique is applied to a conventional PID force/position parallel control scheme which is composed of a PD action on position loop and a proposed adaptive I (integral) action on the force loop, which allows a complete use of available sensor measurements by operating the control action in a full dimensional space without using selection matrices. Simulation results for a three degrees-of-freedom robot show that highly robust position/force tracking can be achieved in the presence of a full dynamic robot and large environment stiffness uncertainties.