Solid oxide fuel cell (SOFC) integrated into micro gas turbine (MGT) cycle is a promising power-generation technology. This article proposes a modified outputinput feedback (OIF) Elman neural network model to describe the nonlinear temperature and power dynamic properties of the SOFC/MGT hybrid system. A physics-based mathematical model of a 220?kW SOFC/MGT hybrid power system is used to generate the data required for the training and prediction of the modified OIF Elman neural network identification model. Compared with the conventional Elman neural network, the simulation results show that the modified OIF Elman identification model can follow the temperature and power response of the SOFC/MGT hybrid system with higher prediction accuracy and faster convergent speed. Copyright (C) 2010 John Wiley & Sons, Ltd.