DNB limit estimation using an adaptive fuzzy inference system

The onset of nucleate boiling is characterized by extremely high heat transfer rates. However, if the fuel rod is operated at a high enough power density, the heat transfer mechanism becomes film boiling with severely reduced heat transfer ability, which is called departure from nucleate boiling (DNB). In this work, the DNB is predicted by an adaptive fuzzy inference system using the measured signals of the average temperature, pressure, and coolant flowrate of a reactor core. An adaptive fuzzy inference system is a fuzzy inference system equipped with a training algorithm. The training method of the adaptive fuzzy inference system is accomplished by two steps: the combined genetic and least-squares algorithms (first step), and the combined backpropagation and least-squares algorithms (second step). The proposed method was verified by using the nuclear and thermal data of the Yonggwang 3 and 4 nuclear power plants. Even though the rule number of this algorithm is small (4 rules), the estimate is accurate. Therefore, this algorithm can provide good information for nuclear power plant operation and diagnosis by predicting the DNB each time step.