Multistability analysis of switched fractional-order recurrent neural networks with time-varying delay

This paper focuses on the existence and dynamic behaviors of multiple equilibria of switched fractional-order recurrent neural networks (SFRNNs) with time-varying delay. By applying the characteristics of Caputo fractional calculus and an effective state space partition method, sufficient criteria are derived to ascertain that the n-neuron SFRNN has 5(n) equilibria, among which 3(n) ones are locally asymptotically stable and the rest are unstable. Moreover, the multistability of integerorder neural networks as a special case is also taken into consideration in this paper. Two illustrative examples are provided to substantiate the theoretical results.