Direct adaptive control based on neural networks

A direct adaptive control scheme based on the fuzzy neural networks (FNN) is presented to diminish the chattering of the sliding mode controller (SMC). In the control scheme SMC and FNN are incorporated by the smooth transformation. We plan to let the SMC force the state to slide into the boundary layer, then use the FNN in the boundary layer to diminish the chattering and obtain smaller tracking error since the FNN has the ability of approximation. The stability of the adaptive control system is proven. Simulation results are given to demonstrate the advantages of the proposed adaptive control scheme.