Event‐Triggered Robust Extended Kalman Filtering for Stochastic Nonlinear Systems Subject to Packet Loss

ABSTRACT In this paper, the event‐triggered robust extended Kalman filter (REKF) design problem for stochastic nonlinear systems subject to packet loss is investigated. Firstly, an event‐triggered REKF with the prediction‐correction structure is constructed to derive the estimation error expression considering packet loss. Secondly, the REKF parameter is obtained by optimizing the estimation error covariance matrix upper bound. The performance of REKF is analyzed and the relationship among the perturbation attenuation performance index, the successful packet transmission probability and the event trigger threshold is obtained. Thirdly, sufficient conditions for the exponential boundedness of the REKF estimation error have been established. Finally, two simulation examples illustrate that the proposed REKF is effective.