Robust adaptive tracking controller design for non-affine non-linear systems with state time-varying delay and unknown dead-zone

A novel Neural Network (NN)-based dead-zone compensation scheme for a class of non-affine Multiple-Input Multiple-Output (MIMO) non-linear systems with state time-varying delay is presented. A static NN is introduced to approximate and adaptively cancel the unknown nonlinear dynamic of the subsystems and unknown dead-zone after its existence proved by implicit function theorem. The control law and adaptive laws for the weights of the hidden layer and output layer of NNs are established by Lyapunov analysis of the whole closed-loop system. The tracking error is proved to be Uniformly Ultimately Bounded (UUB) via introducing an additional robust control term. The effectiveness of the proposed control scheme is verified by the simulation of an MIMO non-affine non-linear system with state time-varying delay.