Artificial Neural Network-Based Nonlinear Black-Box Modeling of Synchronous Generators

This article deals with the black-box modeling of synchronous generators based on artificial neural networks (ANN). The ANN is applied to define relationship between excitation and terminal generator voltage values, Levenberg–Marquardt algorithm used for determining weight coefficients. relation made response reference step changes. proposed approach checked using experimental results obtained from measurements a real 120 MVA hydroelectric power plant Piva in Montenegro. Furthermore, fair comparison nonlinear autoregression model exogenous input (NARX) Hammerstein–Wiener made. For validation, different experiments were conducted—different values disturbances, other controller parameters, rotating speeds. Based presented results, it can be noted that very accurate provides high degree matching outperforms considered models. test procedure are easy implement do not require disconnection grid or additional equipment realization. Such models testing types related system.