Trends in neuro-adaptive control for robot manipulators

An attempt is made to present a method for the adaptive control of nonlinear systems based on a feedforward neural network. The approach incorporates a neurocontroller used within a reinforcement learning framework, which reduces the problem to one of learning an stochastic approximation of an unknown average error surface. Emphasis is placed on the fact that the neurocontroller does not need any input/output information about the controlled system. The proposed method promises to be an efficient tool for adaptive control for both static and dynamic nonlinear systems. Several examples are included to illustrate the scheme