Method and device for detecting defects of metal shaft parts

By preprocessing and feature extraction of ultrasonic data of metal shaft parts, and using machine learning models for defect detection, the problem of low detection accuracy in traditional methods is solved, and high-precision defect detection of metal shaft parts is achieved.

CN120009404BActive Publication Date: 2025-11-18INSTR TECH & ECONOMY INST P R CHINA
2 Cites 0 Cited by

Patent Information

Application Number
CN202510497621.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-04-21
Publication Date
2025-11-18
Estimated Expiration
2045-04-21

AI Technical Summary

Technical Problem

Existing technologies have a high error detection rate in the inspection of metal shaft parts. Traditional methods such as manual visual inspection and magnetic flux leakage detection are affected by subjective factors and are difficult to achieve high-precision defect detection of metal shaft parts.

Method used

By acquiring and preprocessing the raw ultrasonic data of metal shaft parts, key time-domain and frequency-domain features are extracted, and a pre-built metal defect identification model is used for defect detection. Combined with machine learning technology, automated and intelligent defect identification is achieved.

Benefits of technology

It improves the accuracy and recognition rate of defect detection in metal shaft parts, and can identify minute damage that is difficult to detect by traditional methods, thus realizing the automation and intelligence of non-destructive testing.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application relates to the technical field of metal defect detection, in particular to a metal shaft part defect detection method and device. The method comprises the following steps: obtaining original ultrasonic data of a metal shaft part and performing pretreatment to obtain pretreated ultrasonic data, wherein the original ultrasonic data comprises different detection positions and corresponding multi-channel detection data of different detection positions at different angles; performing feature extraction on the pretreated ultrasonic data to obtain key time domain features and key frequency domain features; and inputting the key time domain features and the key frequency domain features into a pre-constructed metal defect recognition model to obtain a defect detection result. The method solves the problem that subtle damage cannot be detected by traditional methods, and improves the defect detection precision and recognition rate.
Need to check novelty before this filing date? Find Prior Art