Stability analysis of reference compensation technique of neural network control for robot manipulators

This paper presents the stability analysis of the reference compensation technique (RCT) of a neural network control scheme (NN) for controlling position of robot manipulators. RCT is known as one of neural network control schemes that compensate for the uncertainties of a robot manipulator. Different from the feedback error learning (FEL) scheme, RCT compensates at the trajectory level so that it has several structural advantages. RCT has been used in many applications. However, the stability analysis of the RCT scheme has not been addressed in the literature. Here the stability is analyzed based on Lyapunov function.