Reachable set bounding for delayed memristive neural networks via adaptive control

This article is concerned with the reachable set estimation (RSE) for delayed memristive neural networks (MNNs). By exploiting the differential inclusion theory and inequality techniques, the RSE problem of MNNs was investigated. A memoryless adaptive controller was designed to realize that states of MNNs converge to a bounded region. Based on this result, an updated memoryless adaptive controller was designed, which further removed the restriction that the delay derivative must be less than 1, leading to a more general result. The new results were presented in the form of algebraic criteria, which were straightforward to verify. Ultimately, the effectiveness of the proposed criteria was demonstrated through two numerical simulations.