Memory-based event-triggered secure state estimation of cyber-physical systems

Abstract

This paper studies the secure state estimation issue of time-varying nonlinear cyber-physical systems under false data injection attacks and a memory event-triggered mechanism. Compared to the conventional event-triggered mechanism based on current system information, a memory-based event-triggered mechanism involving the past measurement output is proposed. Under this scheme, the false triggering induced by abrupt variations of system output can be reduced effectively, and the waste of network resources is mitigated. A Bernoulli variable is utilized to describe the stochastic process of the false data injection attacks. By constructing a novel Lyapunov-Krasovskii functional related to Legendre polynomials and applying Bessel-Legendre inequality technique, sufficient conditions for guaranteeing the asymptotic stability and designing a state estimator of the estimation error system are obtained. Finally, numerical simulations are carried out to illustrate the validity of the presented approach.