Improved synchronisation results for complex-valued neural networks subject to proportional delay: an adaptive intermittent control strategy

This paper is devoted to the synchronisation issue of complex-valued neural networks subject to proportional delay. With the help of Lyapunov stability theory as well as several inequality methods, two types of intermittent controllers can be designed, specifically employing intermittent control and adaptive intermittent control methods, to enable the addressed drive-response systems realise the desired synchronisation. Besides, complex-valued neural networks, which act as a generalisation of real-valued neural networks, possess more complicated properties than the former. Nevertheless, to avoid the need to consider boundedness and analyticity issues, the considered complex-valued neural networks can be effectively decomposed into two practical real-valued subnetworks via the real-virtual separation method, and synchronisation criteria for the two systems are thereby derived. In an effort to validate the efficacy of the above distinct intermittent controllers, in the final simulation experiment, a comprehensive set of visualisation results have been presented. Meanwhile, the inherent advantages of designed adaptive control scheme are also elucidated.