Continuous ZNN Models for Computation of Time-Varying Eigenvalues and Corresponding Eigenvectors

Computation of eigenvalues as well as corresponding eigenvectors of a time-varying matrix (i.e., with time-dependent elements) is attractive. A continuous Zhang neural network (ZNN) model is developed for computing the eigenvalues and corresponding eigenvectors of a time-varying real symmetric matrix. The global convergence to the theoretical eigenvalues and corresponding eigenvectors is proved by theoretical analysis. Meanwhile, the robustness analysis is also presented. In addition, the ZNN model is a dynamic system in essence, which indicates the potential practical implementations and applications. Moreover, numerical experiments are conducted in this paper, and the results verify the effectiveness of the ZNN model. Besides, an extensional ZNN model is discussed in the Appendix.