Damage identification method based on composite material damage physical mechanism orientation using acoustic emission signals

By combining a method guided by the physical mechanism of composite material damage, using acoustic emission signals for segmented processing and a semi-supervised deep neural network classification model, the problem of traditional methods failing to fully consider the underlying physical mechanism is solved, thus achieving accurate identification and comprehensive understanding of composite material damage.

CN117929544BActive Publication Date: 2026-07-24NORTHWESTERN POLYTECHNICAL UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
NORTHWESTERN POLYTECHNICAL UNIV
Filing Date
2024-01-15
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Traditional acoustic emission signal identification methods fail to fully consider the underlying physical mechanisms in composite material damage identification, resulting in poor clustering performance and difficulty in fully understanding and analyzing the damage mechanism.

Method used

A damage identification method based on the physical mechanism of composite material damage is adopted. The acoustic emission signal is segmented and processed. A semi-supervised deep neural network classification model and K-means clustering algorithm are combined. The model is iteratively adjusted to enhance cross-stage learning ability and identify different damage types in composite materials.

Benefits of technology

It improves the accuracy and reliability of damage identification in composite materials, provides a comprehensive understanding of the damage mechanism, and enhances the reliability and accuracy of the identification results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117929544B_ABST
    Figure CN117929544B_ABST
Patent Text Reader

Abstract

The application proposes a damage identification method based on composite material damage physical mechanism guidance using acoustic emission signals, and the acoustic emission data in the composite material is segmented; it is assumed that the acoustic emission data in the first stage is related to the matrix damage, so as to establish a semi-supervised guided deep neural network classification model in the first stage to predict the matrix damage in the second stage data; the clustering algorithm is applied to the data in the second stage which is not marked by the classification model; the data is marked by combining the classification and clustering results, and the semi-supervised guided deep neural network classification model in the second stage is updated; the cross-stage learning ability of the model is enhanced by iteratively adjusting the classification model, and the semi-supervised guided deep neural network classification model is formed. Update all acoustic emission data labels to provide comprehensive damage identification. The application proposes a classification+clustering acoustic emission signal identification method for the physical guidance of composite material damage, which realizes accurate identification of acoustic emission data of composite material damage.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • CN112924556A

  • CN116990119A