Coexistence and locally exponential stability of multiple equilibrium points for fractional-order impulsive control Cohen-Grossberg neural networks
Jinsen Zhang,Xiaobing Nie
DOI: https://doi.org/10.1016/j.neucom.2024.127705
IF: 6
2024-04-19
Neurocomputing
Abstract:Different from the existing multiple Mittag-Leffler stability or multiple asymptotic stability, the multiple exponential stability, which has explicit and faster convergence rate, is investigated in this paper for fractional-order impulsive control Cohen-Grossberg neural networks. First, by using the definition of Dirac delta function, the fractional-order control Cohen-Grossberg neural networks are translated into fractional-order impulsive neural networks, in which pulse effect relies on the fractional order of the addressed system. Then, based on maximum norm, 1− norm and general q− norm (q=2n) , a series of novel criteria are obtained respectively to ensure that such n− neuron neural networks can have ∏i=1n(2Li+1) total equilibrium points and ∏i=1n(Li+1) locally exponentially stable equilibrium points, by utilizing the known fixed point theorem, the method of average impulsive interval, the theory of fractional-order differential equations, and the method of Lyapunov function. This paper's investigation reveals the effects of impulsive function, impulsive interval, and fractional order on the dynamic behaviors. Finally, theoretical results are shown to be effective by four illustrative examples.
computer science, artificial intelligence