Application of Chebyshev neural network to solve Van der Pol equations

In dynamics, the Van der Pol oscillator is a non-conservative with non-linear damping. The problems of single-well, double-well and double-hump Pol-Dufing equations are studied in this paper. Chebyshev Neural Network (ChNN) model will be applied to obtain numerical solutions these types for first time. hidden layer eliminated by expanding input pattern polynomials which employs single neural network. order modify network parameters minimize computed error function, feed forward back propagation principle used. obtained results form ChNN compared analytical solutions, namely Homotopy Perturbation Method (HPM), Analysis (HAM), Differential Transform (DTM) exact. Comparisons existing show that method capable tool solving kind nonlinear problems.