CN-122017045-A - Ultrasonic imaging-based steel member welding correction crack detection method and system

Abstract

The invention relates to the technical field of ultrasonic detection, in particular to a method and a system for detecting welding correction cracks of a steel member based on ultrasonic imaging. The method comprises the steps of obtaining an original ultrasonic signal, performing frequency domain decomposition to obtain frequency domain feature data, performing frequency extraction, threshold comparison and denoising processing on the frequency domain feature data to obtain a denoising ultrasonic signal, extracting a waveform peak value based on the denoising ultrasonic signal to obtain a waveform peak value sequence, performing defect position calculation on the waveform peak value sequence to obtain a defect region coordinate, performing signal interception based on the defect region coordinate to obtain local echo data, performing gradient calculation and peak value extraction according to the local echo data to obtain a crack coordinate position, performing interpolation based on the crack coordinate position to obtain a crack boundary point set, performing outlier rejection based on the crack boundary point set to obtain a stable supporting node, and calculating a closed boundary based on the stable supporting node to obtain a final crack position.

Inventors

• ZHOU YIN

Assignees

• 浙江鸿翔筑能钢构有限公司

Dates

Publication Date

20260512

Application Date

20260124

Claims (8)

1. 1. The method for detecting the welding correction cracks of the steel member based on ultrasonic imaging is characterized by comprising the following steps of: acquiring an original ultrasonic signal, and performing frequency domain decomposition on the original ultrasonic signal to obtain frequency domain characteristic data; extracting components with frequencies exceeding a preset frequency threshold from the frequency domain characteristic data to obtain high-frequency components, and performing threshold comparison and denoising processing based on the high-frequency components to obtain a denoising ultrasonic signal; Extracting waveform peak values based on the denoising ultrasonic signals to obtain waveform peak value sequences, and carrying out defect position calculation on the waveform peak value sequences to obtain defect region coordinates; Intercepting signals based on the coordinates of the defect area to obtain local echo data; Performing gradient calculation and peak value extraction according to the local echo data to obtain a crack coordinate position, and performing interpolation based on the crack coordinate position to obtain a crack boundary point set; and removing abnormal values based on the crack boundary point set to obtain stable supporting nodes, and calculating a closed boundary based on the stable supporting nodes to obtain the final crack position.
2. 2. The ultrasonic imaging-based steel member welding correction crack detection method of claim 1, wherein the acquiring the original ultrasonic signal, performing frequency domain decomposition on the original ultrasonic signal to obtain frequency domain feature data, comprises: Acquiring an original ultrasonic signal, and performing wavelet transformation on the original ultrasonic signal to obtain a wavelet coefficient matrix; And carrying out soft threshold screening according to the wavelet coefficient matrix to obtain a pure wavelet coefficient matrix, and carrying out square operation on the pure wavelet coefficient matrix to obtain frequency domain characteristic data.
3. 3. The ultrasonic imaging-based steel member welding correction crack detection method according to claim 1, wherein the extracting the component with the frequency exceeding the preset frequency threshold from the frequency domain feature data to obtain a high frequency component, and performing threshold comparison and denoising processing based on the high frequency component to obtain a denoising ultrasonic signal comprises: Extracting components with frequencies exceeding a preset frequency threshold according to the frequency domain characteristic data to obtain high-frequency components, and executing Hilbert transformation on the high-frequency components to obtain an instantaneous amplitude envelope; Extracting an area exceeding a preset amplitude threshold value based on the instantaneous amplitude envelope to obtain an interference area set; and carrying out inverse wavelet reconstruction based on the denoising high-frequency component to obtain a denoising ultrasonic signal.
4. 4. The ultrasonic imaging-based steel member welding correction crack detection method according to claim 1, wherein the extracting waveform peak values based on the denoising ultrasonic signals to obtain a waveform peak value sequence, performing defect position calculation on the waveform peak value sequence to obtain defect region coordinates, comprises: extracting a waveform peak value based on the denoising ultrasonic signal to obtain a waveform peak value sequence; Classifying according to the feature vector set to obtain a defect feature vector, calculating a defect coordinate position according to the defect feature vector to obtain a defect position set, and performing space aggregation according to the defect position set to obtain a defect region coordinate.
5. 5. The ultrasonic imaging-based steel member welding correction crack detection method according to claim 4, wherein the signal interception based on the defect area coordinates to obtain local echo data comprises: carrying out signal interception from the denoising ultrasonic signal according to the defect region coordinates to obtain a local signal segment; and windowing the local signal segment to obtain local echo data.
6. 6. The ultrasonic imaging-based steel member welding correction crack detection method according to claim 1, wherein the performing gradient calculation and peak extraction according to the local echo data to obtain a crack coordinate position, and performing interpolation based on the crack coordinate position to obtain a crack boundary point set comprises: performing matrix construction according to the local echo data to obtain an echo intensity matrix, and performing gradient calculation based on the echo intensity matrix to obtain intensity gradient distribution; And carrying out peak extraction according to the intensity gradient distribution to obtain a crack coordinate position, and carrying out linear interpolation according to the crack coordinate position to obtain a crack boundary point set.
7. 7. The ultrasonic imaging-based steel member welding correction crack detection method of claim 1, wherein the performing outlier rejection based on the crack boundary point set to obtain a stable support node, calculating a closed boundary based on the stable support node, and obtaining a final crack position comprises: Calculating curvature according to the crack boundary point set to obtain a curvature distribution set, and eliminating abnormal values based on the curvature distribution set to obtain stable support nodes; And performing polynomial function fitting according to the stable support nodes to obtain a final boundary point sequence, and calculating a closed boundary based on the final boundary point sequence to obtain a final crack position.
8. 8. An ultrasonic imaging-based steel member welding corrective crack detection system, comprising: the frequency domain decomposition module is used for obtaining an original ultrasonic signal, and performing frequency domain decomposition on the original ultrasonic signal to obtain frequency domain characteristic data; The signal denoising module is used for extracting components with frequencies exceeding a preset frequency threshold value from the frequency domain characteristic data to obtain high-frequency components, and performing threshold value comparison and denoising processing based on the high-frequency components to obtain a denoising ultrasonic signal; The defect positioning module is used for extracting waveform peaks based on the denoising ultrasonic signals to obtain waveform peak value sequences, and performing defect position calculation on the waveform peak value sequences to obtain defect region coordinates; The signal interception module is used for intercepting signals based on the coordinates of the defect area to obtain local echo data; the boundary calculation module is used for carrying out gradient calculation and peak value extraction according to the local echo data to obtain a crack coordinate position, and carrying out interpolation based on the crack coordinate position to obtain a crack boundary point set; And the result optimization module is used for removing abnormal values based on the crack boundary point set to obtain stable support nodes, and calculating a closed boundary based on the stable support nodes to obtain the final crack position.

Description

Ultrasonic imaging-based steel member welding correction crack detection method and system Technical Field The invention relates to the technical field of ultrasonic detection, in particular to a method and a system for detecting welding correction cracks of a steel member based on ultrasonic imaging. Background In the field of modern industrial manufacturing and construction, along with the expansion of the application scale of a steel structure, how to effectively identify crack defects of a welding part by using an intelligent sensor and improve the accuracy level of detection becomes a core problem to be solved in the development of the current steel member quality detection technology. In the prior art, the detection of welding cracks mainly depends on the detection mode of ultrasonic flaw detection, and by monitoring the amplitude, time delay or frequency characteristics of ultrasonic echo signals, when a certain index is detected to be out of a preset range, the defect is judged to exist, so that the assessment of welding quality is realized. However, the signal interference environment of the industrial field is increasingly complex, and the fixed signal analysis method is difficult to adapt to the actual defect characteristics under different welding processes and material structures, so that the misjudgment rate is increased or the detection omission risk is increased. In addition, the traditional scheme is often interfered by the non-uniformity in the material and the complex geometric shape of the welding area in the signal propagation process, and the ultrasonic waves can be subjected to multiple scattering and attenuation, so that the judgment of the defect position becomes unclear, and inaccurate crack boundary identification can be caused by the signal superposition effect. In conclusion, the prior art lacks of the fine processing capability of ultrasonic signals in a complex environment and the accurate restoring capability of crack boundary characteristics, so that the problem of insufficient positioning accuracy of welding cracks is caused. Disclosure of Invention The invention provides a method and a system for detecting welding correction cracks of a steel member based on ultrasonic imaging, which can accurately position welding cracks and restore crack boundaries, and solve the problem of insufficient positioning accuracy of the welding cracks in the prior art. In order to solve the technical problems, the invention provides a method for detecting welding correction cracks of a steel member based on ultrasonic imaging, which comprises the following steps: acquiring an original ultrasonic signal, and performing frequency domain decomposition on the original ultrasonic signal to obtain frequency domain characteristic data; extracting components with frequencies exceeding a preset frequency threshold from the frequency domain characteristic data to obtain high-frequency components, and performing threshold comparison and denoising processing based on the high-frequency components to obtain a denoising ultrasonic signal; Extracting waveform peak values based on the denoising ultrasonic signals to obtain waveform peak value sequences, and carrying out defect position calculation on the waveform peak value sequences to obtain defect region coordinates; Intercepting signals based on the coordinates of the defect area to obtain local echo data; Performing gradient calculation and peak value extraction according to the local echo data to obtain a crack coordinate position, and performing interpolation based on the crack coordinate position to obtain a crack boundary point set; and removing abnormal values based on the crack boundary point set to obtain stable supporting nodes, and calculating a closed boundary based on the stable supporting nodes to obtain the final crack position. In an optional implementation manner, the acquiring the original ultrasonic signal, performing frequency domain decomposition on the original ultrasonic signal to obtain frequency domain feature data, includes: Acquiring an original ultrasonic signal, and performing wavelet transformation on the original ultrasonic signal to obtain a wavelet coefficient matrix; And carrying out soft threshold screening according to the wavelet coefficient matrix to obtain a pure wavelet coefficient matrix, and carrying out square operation on the pure wavelet coefficient matrix to obtain frequency domain characteristic data. In an optional implementation manner, the extracting the component with the frequency exceeding the preset frequency threshold from the frequency domain feature data to obtain a high frequency component, performing threshold comparison and denoising processing based on the high frequency component to obtain a denoised ultrasonic signal, and includes: Extracting components with frequencies exceeding a preset frequency threshold according to the frequency domain characteristic data to obtain high-frequency componen