时滞切换神经网络系统的鲁棒指数稳定性

Robust Exponential Stability of Switched Neural Networks Systems with Time Delay

研究了具有时间滞后切换不确定细胞神经网络(UCNNs)系统的指数稳定性.利用同胚映射和M矩阵理论,得到UCNNs系统平衡点存在性,唯一性和指数稳定性的充分条件;利用平均驻留时间方法,研究了时滞切换UCNNs系统限制切换下的鲁棒指数稳定性,并得到确保系统全局指数稳定的充分条件.

In this paper, a class of switched uncertain cellular neural networks （UCNNs） systems with unbounded delay were investigated. By using the homeomorphic mapping theorem and M-matrix theory, sufficient conditions for the existence, uniqueness and ex- ponential stability of the equilibrium point of UCNNs systems with unbounded delay were obtained. By using average dwell time approach, we studied the robust exponential stability of constrained switching delay UCNNs systems. Sufficient conditions to guarantee the global exponential stability of switched delay UCNNs systems were derived and the stability was robust to both parameter perturbations and switching perturbations.