Small-Gain-Based Adaptive Secure Control of Switched Cyber-Physical Systems With Denial-of-Service Attacks

The article investigates the secure control against denial-of-service (DoS) attacks in sensor-controller channel and actuator fault for switched cyberphysical systems (CPSs). First, the concerned CPSs are subjected to the switching features and strict-feedback nonlinearities, and hence a compound switched system is exposited. Second, a switched state observer is formulated, while neural networks are introduced to identify the nonlinear functions and offset the actuator faults. Third, a switched controller is constructed based on the estimated output or measured output, depending on the DoS attack is active or sleeping. The resulted switched closed-loop system is proved to be input-to-state practically stable by the small-gain theorem, average dwell time, common Lyapunov function, and multiple Lyapunov function. Finally, numerical simulation validates the effectiveness of the suggested control.