Stability Prediction of Nonlinear System Using Multilayer Feed-forward Artificial Neural Network

Stability analysis of linear systems can be studied by root-locus technique, Nyquist criterion and Routh Hurwitz's stability testing. Stability study of nonlinear systems can be done by describing function method, phase-plane analysis, numerical integration, and Lyapunov's second method. In this paper, neural network approach to predict the stability of nonlinear systems is presented. This approach is illustrated with an example.