Nonlinear dynamic system identification using neuro

This paper presents a neuro-fractional-order Hammerstein model with a systematic identification algorithm for identifying unknown nonlinear dynamic systems. The proposed model consists of a Radial Basis Function Neural Network (RBF NN) followed by a Fractional-Order System (FOS). The proposed identification scheme is performed in two stages. First, the fractional-order and the number of state variables (or degree) of the state-space realization of the FOS are estimated in the frequency domain. Then, the parameters of the RBF NN (the weights, centers and widths of the Gaussian functions) and the state matrix of the FOS are determined using the time domain data via the Lyapunov stability theory. Simulating as well as experimental examples are provided to verify the effectiveness of the proposed method. The identification results show that the proposed neuro-fractional-order Hammerstein modeling is superior as compared with the existing Hammerstein modeling in literature