Neural network based compensator for robustness to the robot manipulators with uncertainties

In this paper, neural network based compensator is developed to estimate the bound of structured and unstructured uncertainties in the robot dynamics to provide a adaptive robust controller. Especially, the prior knowledge of the upper bound of the system uncertainties is not required for designing of tracking controller. Lyapunov approach will be used to show that the filtered tracking error and neural network weight error are uniformly ultimately bounded. It is found that the feedforward neural network is effectively able to cope with all uncertainties existing in the robot manipulator. Finally, simulation studies are carried out for a two-link robot manipulator to show the effectiveness of the control scheme.