Stability analysis of neural networks with time delays via energy functions approach

In this paper, we improve the approach of energy functions on the stability analysis of neural networks with time delays. By means of coefficient variation as well as inequality analysis, a set of stability criteria for neural networks with time delays are given and the relationship between equilibrium points of the network and local minimum points of the energy function is discussed. As an application, we provide a neural network model that can be applied to calculate all the eigenvectors with respect to the maximum eigenvalue of a real symmetric matrix.