Adaptive Global Secure Control for Nonlinear Cyber–Physical Systems Under Deception Attacks

This paper studies the problem of adaptive global secure control for nonlinear cyber-physical systems (CPSs) encountering unknown deception attacks. First, the radial basis function neural network (RBFNN) is adopted to approximate the unknown functions of the system, while a nonlinear function is designed to dominate the uncertain dynamics of the system. Then, by leveraging the computational advantage of the single-parameter estimation methodology, an auxiliary variable is constructed to mitigate the impact of deception attacks, and an improved secure controller is designed to guarantee the global boundedness of all closed-loop signals. In contrast to the existing approaches that typically ensure only local or semi-global boundedness, the proposed control scheme not only achieves the global boundedness of all system signals in the presence of unknown deception attacks but also eliminates the restrictive assumption that attack signals are strictly positive. Finally, the simulation results with an RLC circuit system demonstrate the effectiveness of the proposed approach.