A penalty-based adaptive secure estimation for power systems under false data injection attacks

Injection attack estimation of networked control systems subject to hidden DoS attack

This paper is concerned with the injection attack estimation for a class of networked control systems with Deny-of-Service (DoS) attack, where unknown signals are injected into the sensor reading and actuator. The main goal is to design an estimator to estimate the injected signals subject to stochastic hidden DoS attack. By introducing a semi-Markov chain, the complex aperiodic sampling behavior caused by DoS attack is first modeled as a stochastic switching system. By using the Lyapunov stability theory and stochastic system analysis method, the σ-error mean square stability (σ-EMSS) conditions of estimation error system are then derived and the state of the attack estimation error system is shown to be uniformly ultimately bounded. Some linear matrix inequalities are proposed to determine the estimation gain matrices. Finally, both simulation and experimental studies on the motor system are conducted, and the effectiveness of the main results are demonstrated.