Uncertainty Calculation Method for Data Based Model

摘要

The present invention provides an uncertainty calculation method of a data-based model which calculates uncertainty of prediction data of a data-based model monitoring a drift of a sensor used in a nuclear power plant and can increase reliability of the prediction data by the calculated uncertainty of the prediction data. The uncertainty calculation method of a data-based model comprises a memory data generating step (S10), a measured data receiving step (S20), an Euclidean distance calculating step (S30), a kernel function calculating step (S40), a calculating step of an effective number for each weight area (S50), a weight setting step (S60), a total effective number calculating step (S70), a prediction data calculating step (S80), a weight standard deviation calculating step (S90), and an uncertainty calculating step (S100).