This paper proposes a fuzzy predictive control scheme for controlling power output of a boiler turbine system in the presence of disturbances and uncertainties. A new model of the boiler turbine system is introduced based on the modeling approaches of hybrid systems, namely, the mixed logical dynamical modeling approach. Nonlinear parts of the system are linearized using the piecewise affine approach. To overcome the deficiency of the model predictive control in presence of disturbance and uncertainty, a fuzzy predictive control scheme is proposed in which a fuzzy supervisor is utilized to adjust the main predictive controller. The proposed fuzzy predictive control scheme has advantages such as simplicity and efficiency in nominal conditions and strong robustness in the presence of disturbances and uncertainties. Simulation results demonstrate the effectiveness and superiority of the method.