Finite/fixed-time synchronization of complex-valued switched coupled neural networks via unified control strategy.

Abstract

Since complex-valued neural networks (CVNNs) offer inherent advantages over real-valued neural networks in representing intricate information and handling multi-dimensional data, the dynamic behavior of CVNNs with the switched topologies is analyzed. The finite-time synchronization (FTS) and fixed-time synchronization (FXTS) of complex-valued switched coupled neural networks (CVSCNNs) are investigated in this article via the unified control strategy. Existing results regarding the synchronization of complex-valued switched coupled neural networks are relatively limited and predominantly rely on the single control strategy. Consequently, the realization and analysis of synchronization are difficult to perform in neural networks with switched topologies and coupled relationships, which urgently requires the proposal of a unified control strategy. Firstly, by considering the comprehensive impact of switched topologies and coupled relationships in CVSCNNs, several conditions related to FXTS are derived via the non-separation methodology, and the estimation of corresponding settling time (ST) is obtained. Subsequently, CVSCNNs are transformed into equivalent real-valued systems by decomposition methods, and the FTS and FXTS of CVSCNNs are analyzed by the unified control strategy. Furthermore, compared with the results of the existing single control strategy, the FTS and FXTS of the drive and response systems are achieved via unified control strategy, avoiding the need for separate controllers in different scenarios. Finally, the effectiveness of the obtained results is further verified by two numerical examples.