Direct adaptive output feedback control for MIMO uncertain nonlinear discrete-time systems based on neural networks

This paper develops a novel neural-network-based direct adaptive control scheme for a class of multi-input-multi-output uncertain nonlinear discrete-time (DT) systems in the presence of unknown bounded disturbances. By employing feedback linearization methods, neural network (NN) approximation can cancel the nonlinearity of the DT systems. Meanwhile, the weights of NNs are directly updated online instead of preliminary offline training. In addition, unlike most literatures, the condition for persistent excitation is removed. Based on Lyapunov´s direct method, both tracking errors and weight estimates are guaranteed to be uniformly ultimately bounded, while keeping the closed-loop system stable. Finally, an example is provided to demonstrate the effectiveness of the proposed approach.