Adaptive neural control for a class of uncertain chaotic system with non-affine input

A robust adaptive neural network control method is proposed for a class of chaotic systems with non-affine inputs and uncertain disturbances. The assumptions that the non-affine term is differentiable with respect to input and the partial derivative must be positive are cancelled in our proposed method. The adaptive compensation term is adopted to minify the influence of approximation error and uncertain disturbances. By using the Lyapunov function, it was demonstrated that all signals involved are bounded, and the tracking error converges to a small neighborhood of origin. The proposed scheme is applied to the Duffing-Holmes and Genesio chaotic systems, simulation results demonstrate the effectiveness of this method.