Fixed-Time Control for Memristor-Based Quaternion-Valued Neural Networks with Discontinuous Activation Functions.

Ruoyu Wei,Jinde Cao,Sergey Gorbachev
DOI: https://doi.org/10.1007/s12559-022-10057-9
IF: 4.89
2022-01-01
Cognitive Computation
Abstract:This paper studies the fixed-time synchronization control of quaternion-valued memristive neural networks (QVMNNs). The QVMNN is the extension of real and complex-valued MNNs, and the activation functions considered in this work are assumed to be discontinuous. Due to the noncommutativity of quaternion multiplication, the QVMNNs model is separated into four real-valued systems by utilizing the differential inclusion theory and decomposition method. Based on the sign function, some discontinuous control schemes are developed. By applying the nonsmooth analysis and inequality techniques, some novel criteria for fixed-time synchronization of QVMNNs are derived. Compared with the previous results, the proposed method based on sign function makes the designed controllers more concise and the established criteria more effective and less conservative. Finally, simulations are proposed to demonstrate the validity and practicability of theoretical results.
What problem does this paper attempt to address?