Development of adaptive control system using the fuzzy theory for thermal-hydraulic analysis code, AQUA

An adaptive control system using the Fuzzy theory has been implemented into the single-phase three-dimensional thermal-hydraulic analysis code AQUA. The system is designed to yield the optimum time step sizes as a control parameter. It fully utilizes the experiences accumulated by numerical experts: e.g., the use of a very small time step size when a symptom of instabilities appear. Asymtotic behaviors in the course of numerical integration due to AQUA are monitored by checking mass residual and relative variations for three velocity components u, v and w, turbulent parameters k and ε and enthalpy h every time step. The validity of overall idea could be interpreted by analogy with the fact experts in many industries are capable of operating plants without profound knowledge of the control characteristics. As a result of numerical experiments, the total CPU time has been reduced typically by 40% in comparison with the cases when the time step size is not controlled by the system. In general, the adaptive control system using the Fuzzy theory is an efficient measure to save computing efforts when one wishes to perform an extremely large scale simulation over a long transient time span. (author)