Chaos control of Bonhoeffer-van der Pol oscillator using neural networks (Q1600555)

The present paper deals with the control of chaos of a continuous time dynamical systems namely the Bonhoeffer-van der Pol (BVP shortly) oscillator using the feed-forward backpropagating neural network. The neural network is trained on two different controlling algorithms, namely the Ott-Grabogi-Yorke (OGY) method and Pyragas method of delayed feedback control. It is found that in the neural network trained on the OGY method, the control transient are smaller than using the direct OGY algorithm. However, in the neural network trained on the Pyragas method, the control transients are of the same order.