A novel resilient event-triggered secure state estimation approach for power networks

For cyber physical systems such as power networks, event-triggered communication presents an effective strategy to reduce data transmission between devices with limited energy and computational resources. However, the performance of such systems can degrade significantly in the presence of cyberattacks. This paper proposes a secure event-triggered state estimation framework for power networks under sensor attacks. By utilizing the analytical redundancy inherent in the power system, a bank of Kalman filters is utilized to identify attack-free sensors and estimate the system state. Given that Kalman filtering provides minimum mean squared error estimation under the assumption of Gaussian innovations, we offer a formal proof that our design preserves the Gaussianity of the innovation process. To validate the effectiveness of the proposed estimator, we conduct a simulation study on the IEEE 14-bus power network, comparing its performance against a conventional estimator. Our results demonstrate that the average trace of the estimation error covariance achieved by the proposed method is 0.3146, with a communication rate of approximately 36.72%, whereas the average trace of the estimation error covariance for the conventional estimator is 0.4262 and the average communication rate is 79.78%. These statistics emphasize the superior performance of the proposed estimator. • Developed an event-triggered communication to reduce data transmission. • Proposed Kalman filter-based secure estimation to enhance CPS cybersecurity. • Utilized analytical redundancy for robust sensor attack detection and mitigation. • Designed recursive Kalman filters for efficient anomaly detection in CPSs. • Validated estimator's superiority using IEEE 14-bus benchmark simulations.