Exponential Stability Analysis of Neural Networks with Multiple Time Delays
Without assuming the boundedness, strict monotonicity and differentiability of the activation function, a result is established for the global exponential stability of a class of neural networks with multiple time delays. A new sufficient condition guaranteeing the uniqueness and global exponential stability of the equilibrium point is established. The new stability criterion imposes constraints, expressed by a linear matrix inequality, on the self-feedback connection matrix and interconnection matrices independent of the time delays. The stability criterion is compared with some existing results, and it is found to be less conservative than existing ones.
Unable to display preview. Download preview PDF.