Abstract
In this paper, the switching dynamics of recurrent neural networks are studied. Sufficient conditions on global exponential stability with an arbitrary switching law or a dwell time switching law and the estimates of Lyapunov exponent are obtained. The obtained results can be used to analyze and synthesize a family of continuous-time configurations with the switching between the configurations. Specially, the obtained results are new and efficacious for the switching between the stable and unstable configurations. Finally, simulation results are discussed to illustrate the theoretical results. © 2009 Springer Berlin Heidelberg.
Cite
CITATION STYLE
Zeng, Z., Wang, J., & Huang, T. (2009). Global exponential stability of recurrent neural networks with time-dependent switching dynamics. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5769 LNCS, pp. 583–592). https://doi.org/10.1007/978-3-642-04277-5_59
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
