Integral sliding mode secure control for linear systems subject to malicious attacks: A hybrid learning scheme

In this paper, the secure control problem of linear time-invariant (LTI) systems under malicious false data injection attacks (FDIAs) is investigated. Initially, the optimal control problem of LTI systems is resorted to solving the algebraic Riccati equations (AREs). Subsequently, a hybrid iteration scheme is proposed that combines the advantages of both value iteration and policy iteration to solve the optimal control problem of LTI systems in the case where the system dynamics information is unknown. This scheme not only eliminates the requirement for an initial stabilizing control policy but also has a faster convergence rate compared to value iteration. At the same time, a super-twisting integral sliding mode controller, designed based on the hybrid iteration scheme, is applied to resist malicious FDIAs without the requirement for system dynamics information and reduce the chattering phenomenon. Finally, the effectiveness of the proposed control scheme is validated through an operational amplifier circuit.