An intelligent computational algorithm based on neural networks for the identification of chaotic systems

Abstract

The identification of nonlinear systems with chaotic behavior using a neural network based computational algorithm is presented.. A neural network is trained on the measured output data of the actual system. The network parameters viz. the neural network weights are estimated using the Elman back propagation algorithm .Further, The Rossler and the Chen chaotic systems are used for simulation. The simulation results show that the ANN trained with back propagation algorithm performs very well and give exact reproduction of the output time series and states, as generated from the dynamical equations. The Kaplan Yorke dimensions and the Lyapunov exponents of the model are calculated.