On State Estimation for Discrete Time-Delayed Memristive Neural Networks Under the WTOD Protocol: A Resilient Set-Membership Approach

In this article, a resilient set-membership approach is put forward to deal with the state estimation problem for a sort of discrete-time memristive neural networks (DMNNs) with hybrid time delays under the weighted try-once-discard protocol (WTODP). The WTODP is utilized to mitigate unnecessary network congestion occurring in the channel between DMNNs and the state estimator. In order to ensure resilience against possible realization errors, the estimator gain is permitted to undergo some norm-bounded parameter drifts. Our objective is to design a resilient set-membership estimator (RSME) that is capable of resisting gain variations and unknown-but-bounded noises by confining the estimation error to certain ellipsoidal regions. By resorting to the recursive matrix inequality technique, sufficient conditions are acquired for the existence of the expected RSME and, subsequently, an optimization problem is formalized by minimizing the constraint ellipsoid (with respect to the estimation error) under WTODP. Finally, numerical simulation is carried out to validate the usefulness of RSME.

Automated summary

This study presents a resilient set-membership approach for addressing the state estimation problem of discrete-time memristive neural networks, using the weighted try-once-discard protocol to alleviate network congestion. A resilient set-membership estimator is designed to resist gain variations and unknown-but-bounded noises by confining the estimation error within certain ellipsoidal regions. Sufficient conditions for the existence of the estimator are obtained through recursive matrix inequality technique, and an optimization problem is formalized to minimize the constraint ellipsoid under the weighted try-once-discard protocol.