A method and apparatus for predicting leakage values ​​based on ultrasound.

By using ultrasonic methods in nuclear power plants or nuclear engineering projects, acquiring training parameters through data acquisition devices, calculating mutual information coefficients, filtering feature combination datasets, and constructing an integrated prediction model, the problem of real-time online prediction of leakage values ​​is solved, achieving highly accurate and reliable leakage detection.

CN119665154BActive Publication Date: 2026-05-26CHINA NUCLEAR POWER ENGINEERING CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA NUCLEAR POWER ENGINEERING CO LTD
Filing Date
2024-11-26
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

There is currently no technology that uses ultrasonic methods to predict leakage values ​​in critical areas of nuclear power plants or nuclear engineering projects in real time based on leakage mechanisms and phenomena.

Method used

Training parameters are acquired through a data acquisition device, mutual information coefficients are calculated, relevant feature combination datasets are selected, an integrated prediction model is constructed, and leakage values ​​are predicted using the ultrasonic method.

Benefits of technology

It improves the accuracy and reliability of leak detection, reduces reliance on engineers' experience, and enhances the model's generalization ability and prediction accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119665154B_ABST
    Figure CN119665154B_ABST
Patent Text Reader

Abstract

This invention discloses a method and apparatus for predicting leakage values ​​based on ultrasound. The method includes acquiring training parameters under different leakage conditions using a data acquisition device; calculating a mutual information coefficient based on the training parameters; training an integrated prediction model using the training parameters and the mutual information coefficient; acquiring detection parameters using a pipeline monitoring device; inputting the detection parameters into the integrated prediction model to obtain prediction results; and analyzing the prediction results. This invention provides a method for predicting leakage values ​​based on ultrasound that does not rely on engineer experience or complex process knowledge, and can automatically select suitable input variables to train the model and predict leakage values.
Need to check novelty before this filing date? Find Prior Art