Secure State Estimation of Singularly Perturbed Switched Systems With Output Deception Attack via Discretized LKF Approach.

In this brief, the secure mean-square asymptotic state estimation (SMSASE) of a singularly perturbed switched system (SPSS) with delay and output deception attack is investigated. An attack decomposition method (ADM) is proposed to avoid directly utilizing the attacked output signal, by which a non-attack signal is obtained and utilized to design a full-order estimator such that the state of the original system is precisely estimated. To minimize the influence of switching and time delay on the conservatism of the obtained results, a discretized Lyapunov-Krasovskii functional (LKF) is designed to guarantee the increment coefficient of the LKF at the switching instants to be smaller than one, facilitating easy derivation of non-weighted L2 gain without any additional inequality transformation. The estimation criterion and the design of the estimator gains are provided as linear matrix inequalities (LMIs). The validity of the theoretical analysis is verified through a numerical simulation.