Sensor Fault Estimation of Complex Networks under Event-Triggered Communication

The problem of sensor fault estimation for a kind of nonlinear discrete time-varying complex networks is studied under event-triggered communication. The event-triggered communication protocol is introduced to save communication resources. Two kinds of faults, including the unknown constant fault and the slow drift fault, are considered, which are both modeled as a linear function. For such systems, combining faults with states into augmented states, a distributed estimator for the augmented system is constructed. Then, an estimator gain matrix is obtained to minimize the upper bound which is derived by scaling the estimation error covariance. Finally, a numerical simulation is applied to verify the validity of the estimation method.