Nonlinear systems identification using dynamic multi-time scales neural networks

In this paper, an on-line identification algorithm is proposed for nonlinear systems identification via dynamic neural networks with different time-scales including both fast and slow phenomenon. The main contribution of the paper is that the Lyapunov function and singularly perturbed techniques are used to develop the on-line update laws for both dynamic neural networks weights and the linear part matrices of the neural network model. Compared with the other dynamic neural network identification methods, the proposed identification method exhibits improved identification performance. Two examples are given to demonstrate the effectiveness of the theoretical results.