作者: Zhenkun Huang , S. Mohamad , Xinghua Wang , Chunhua Feng
DOI: 10.1002/CTA.490
关键词:
摘要: In this paper, we investigate convergence dynamics of $2^N$ almost periodic encoded patterns general neural networks (GNNs) subjected to external stimuli, including delays. Invariant regions are established for the existence under two classes activation functions. By employing property $\mathscr{M}$-cone and inequality technique, attracting basins estimated some criteria derived converge exponentially toward patterns. The obtained results new, they extend generalize corresponding existing in previous literature.