Adaptive neural network backstepping predictive control for nonlinear time-delay system

A new adaptive Neural Networks (NN) backstepping feedback control strategy is presented for a class of uncertain affine nonlinear systems, which arc with constant input time-delay and in strict-feedback form. The proposed algorithm consists of a predictive feedback strategy and an adaptive High Order Neural Networks (HONN) backstepping controller. The predictive control mechanism is constituted to compensate the effect of the input time-delay. HONNs are utilized to approximate the integrative unknown nonlinear term of the controller incorporated with a special backstepping control configuration that is local to the controlled plant, thus the controller singularity problem can be removed completely. The output of system is guaranteed to track the delayed reference signal. Rigorous stability analysis for both the adaptive HONN controller and the tracking error of the closed-loop control system is investigated by means of Lyapunov stability criterion. Simulation results are included to demonstrate the validity and simplicity of the proposed controller.