Skip to Main Content
FRIT (fictitious reference iterative tuning) is attractive in the field of the control system design. Especially, the FRIT has some useful features in practical use. One is that it dose not require the system identification. Another is that the control parameters can be directly computed using only the closed-loop input/output data and the desired output signal. In this paper, the FRIT is extended for discrete-time stochastic systems. Then, the real-coded genetic algorithm is employed in optimizing the criterion. The effectiveness of the proposed discrete FRIT is verified by a numerical simulation example and an application for a pilot-scale heat process.