Study on the delayed feedback control of chaos in chaotic neural networks

Abstract

Chaotic neural networks consisting of chaotic neurons exhibit rich dynamic behaviors and are expected to be used in information processing. But the output sequence of chaotic neural networks is chaotic, so the networks do not converge to a stable pattern. In order to apply chaotic neural networks to information search or pattern recognition, etc., it is necessary to control chaos in chaotic neural networks. In this paper, we propose an improved delayed feedback control method for chaotic neural networks. By means of the control method, computer simulation shows that controlled chaotic neural networks can converge to period-2 states between one stored pattern and its reverse pattern or various multiple-period states depending on the delay time.