Gaussian process regression-based critical heat flux density prediction method

Abstract

The invention discloses a Gaussian process regression-based critical heat flux density prediction method, which is specifically implemented according to the following steps of: 1, dividing a collecteddata set into two parts, namely 70% of data as a training set and the balance of 30% of data as a test set, and recording D * = [(X *, y *)}; 2, preprocessing the training set data and the test set data by adopting a standard normal variable method to enable the mean value of the training set D and the test set D * to be 0 and the standard deviation to be 1; 3, inferring the relation between thetraining input variable and the training target output by using Gaussian process regression to obtain a critical heat flux prediction model; And 4, predicting the critical heat flux density through the system pressure P, the mass flow rate G and the balanced steam content Xe by using the obtained critical heat flux density prediction model. The method can accurately and effectively predict the critical heat flux density.

Description

technical field [0001] The invention belongs to the field of reactor core safety analysis, in particular to a critical heat flux prediction method based on Gaussian process regression. Background technique [0002] In the safety review of nuclear reactors, the critical heat flux is an important thermal-hydraulic limiting parameter, which refers to the maximum heat flux that can be tolerated before the temperature of the heated wall surface soars and burns out. Once the heat flux exceeds the critical heat flux, it will cause the wall temperature to overheat, which will cause the element to burn out. Therefore, it is very important to predict the critical heat flux accurately for the safety and economy of the reactor. [0003] Because the critical heat flux is a very complex phenomenon, although more than 500 critical heat flux prediction methods have appeared in the literature in the past few decades, there is still no unified theory that can accurately predict the critical ...

AI Technical Summary

Problems solved by technology

Although the artificial neural network can overcome the shortcomings of the traditional critical heat flux prediction method to a certain extent, because the artificial neural network is a method based on the principle of structural risk minimization, it is difficult to choose the structure, easy to fall into the minimum value, and overfitting. And disadvantages such as poor promotion ability

Images