Composite sliding mode control for cyber-physical systems with multi-source disturbances and false data injection attack

This paper investigates the sliding mode control problem for cyber-physical systems subject to multi-source disturbances and actuator false data injection attacks simultaneously. Firstly, a series of extended state observers are designed to estimate the unknown multi-source disturbances. For the actuator false data injection attacks, a sliding mode observer is designed to online estimate the values of attacks. Then, by introducing the disturbance and attack estimations, a dynamic sliding manifold and a composite sliding mode controller are constructed. Under the proposed approach, the unknown disturbances and false data injection attacks can be timely estimated and compensated, which can improve the disturbance rejection ability and guarantee security of the cyber-physical system. Lyapunov theory is utilized to prove the stability of the closed-loop system. Finally, simulations of both the numerical example and application to a power converter system demonstrate the effectiveness and superiority of the proposed composite sliding mode control approach.