Adaptive synchronization of memristor-based neural networks with discontinuous activations
Neurocomputing(2020)
摘要
This paper investigates the adaptive synchronization of memristor-based neural networks (MNNs) with discontinuous activations. First, by using a decoupling strategy, the extensively used models of MNNs are improved, which can describe the real dynamics more accurately. Moreover, we extend the asymptotic synchronization criteria for MNNs with the consideration of discontinuous activation functions. Additionally, a general adaptive controller is devised to synchronize the drive and response systems, and the sufficient stability conditions are established via Lyapunov functional method within the framework of differential inclusions. Finally, numerical simulations are conducted to show the effectiveness of the developed methods.
更多查看译文
关键词
Memristor-based neural networks,Discontinuous activations,Asymptotic synchronization,Adaptive control
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络