Adaptive neural network finite-time control for nonlinear cyber-physical systems with external disturbances under malicious attacks

Abstract

This paper is concerned with the finite-time tracking control problem for a class of high-order nonlinear cyber-physical systems (CPSs) with external disturbances, which are subject to malicious attacks occurring in controller-actuator (C-A) channel. Combining the variable structure (VS) control method and the artificial neural network (NN) technique, a novel adaptive neural network finite-time control strategy is developed. In particular, a Gaussian radial basis function neural network (RBFNN) is designed as an adaptive neural estimator working in an online manner to achieve the estimation and reconstruction of malicious attacks and external disturbances. Furthermore, rigorous proofs have been presented to show that the proposed control scheme can guarantee that the output tracking error converges to the origin in finite time. Finally, the proposed control scheme is applied to heavy duty vehicle system (HDVS) as a testbed, and a representative simulation is provided to demonstrate the validity and effectiveness of the proposed method.