The development of automation facilities and microcontrollers makes possible implementation and industrial application of complex control algorithms. However, the algorithms, which differ from classical PI / PID control laws, are not widespread. Considering disturbances rejection, the PID algorithm, like any other feedback law, reacts to control error rise, but such transient quality could be improved in case the controller would be able to react earlier. In this research mentioned problem is solved with the help of the proposed predictive controller based on fuzzy logic, which is used to reject technological disturbances acting on such plants as heating furnaces. The process of fuzzy system design is described; the results of comparative modeling are shown. Application of the developed fuzzy predictive controller allows to improve the compensation of such disturbance as inter-zone heat transfer in comparison with a conventional PI-controller. The amplitude of the control error is reduced by 64% averagely comparing to PI-controller tuned for transient processes, the transient time is reduced by 21%.