CN103901115A - Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising - Google Patents
Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising Download PDFInfo
- Publication number
- CN103901115A CN103901115A CN201410150224.0A CN201410150224A CN103901115A CN 103901115 A CN103901115 A CN 103901115A CN 201410150224 A CN201410150224 A CN 201410150224A CN 103901115 A CN103901115 A CN 103901115A
- Authority
- CN
- China
- Prior art keywords
- signal
- emd
- coarse grain
- grain material
- wavelet
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention discloses an ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising. A low-signal-to-noise-ratio signal for detecting a coarse grain material is denoised based on the adaptive decomposition capacity of EMD to a non-steady nonlinear signal and the wavelet multi-scale multi-resolution denoising capacity, so that the signal-to-noise ratio of the detection signal is fully improved, and effective detection on the coarse grain material is realized. The ultrasonic coarse grain material detection method has the advantages that the advantages of the EMD and the wavelet threshold denoising for detection signal processing are fully combined, so that the detection signal with low signal-to-noise ratio can be reasonably denoised, and the signal and noise can be effectively distinguished; the noise is removed by selecting a proper threshold, the signal-to-noise ratio is greatly increased, and the aim of effectively detecting the coarse grain material is finally fulfilled.
Description
Technical field
The present invention is adapted to the defect location of low signal-to-noise ratio detection signal, is specifically related to a kind of coarse grain material supersonic detection method based on EMD associating wavelet threshold denoising.
Background technology
The advantages such as that coarse grain material has is anticorrosive, high temperature creep-resisting ability is strong, in the last few years, in modern industry application more and more extensive, as having purposes widely in the industrial sectors such as chemical reaction container and pipeline, the storage of liquid gas and transportation, nuclear power.But in the time that it is carried out to Ultrasonic Detection, construct noise has seriously reduced the signal to noise ratio (S/N ratio) of detection signal, defect reflection is usually flooded by noise and is difficult to identify.Therefore, no matter in process of production, or in daily servicing, this class material is carried out to Non-Destructive Testing and be absolutely necessary.
Non-Destructive Testing is the detection technique that produces and develop on modern science and technology basis, it is by advanced technology and instrument and equipment, in the situation that not damaging, not changing detected object physics and chemistry state, inside to detected object and surperficial structure, character, state carry out inspection and the test of what high reliability of high sensitivity, so as to passing judgment on their continuity, integrality, security other performance index together.Non-Destructive Testing is as the effective detection means of one, be widely used in the every field of economic construction in China, the manufacture of for example special equipment detects and with check, and the industry such as machinery, metallurgy, petroleum gas, chemical industry, Aero-Space, boats and ships railway, electric power, building.Especially guaranteeing pressure-bearing class special equipment product quality and using secure context, it is particularly important that Dynamic Non-Destruction Measurement seems.Ultrasonic detecting technology has that large, highly sensitive, the penetration power of the degree of depth of detection is strong, accurate positioning, with low cost, speed is fast, harmless and be convenient to the features such as on-the-spot use, therefore, compare with other nondestructiving detecting means, ultrasonic detecting technology is most widely used general.In the time that coarse grain material is carried out to Ultrasonic Detection, random large crystal grain boundary will produce strong scattering to ultrasound wave, cause serious construct noise and ultrasonic energy decay, the various random noises of adding the mixed test macro of acoustic-electric string between transmitting receiving transducer, cause Ultrasonic Detection sensitivity and defect detection rate degradation.How very noisy signal is carried out to squelch, improving signal to noise ratio (S/N ratio) and defect detection rate is the emphasis of this research work.
The people such as N.E. Huang in 1998 have proposed empirical mode decomposition method (EMD), be applicable to analyze the nonlinear time-varying process of non-stationary, the difference of itself and wavelet analysis is that it is posterior, do not need to select in advance basis function, but producing adaptively suitable mode function according to the characteristic of signal itself, these mode functions are the reflected signal frequency characteristic of part at any time well.Because it can resolve into sophisticated signal a series of frequencies intrinsic mode function (IMF) from high to low, therefore can be seen as pickup electrode value tag yardstick is the spatio-temporal filtering process of tolerance, and can utilize this character to carry out Filtering Analysis and denoising to signal.But utilize EMD to complete filtering and noise reduction and just deduct several IMF components after decomposition with original signal simply, cause the useful information filtering together with noise in respective component, thereby cause distorted signals.
Wavelet analysis is a frontier developing rapidly in current application mathematics and engineering discipline, and wavelet transformation has contacted multiple subjects such as applied mathematics, physics, computer science, Signal and Information Processing, image processing, seismic prospecting.Wavelet analysis is widely applied in signal process field in recent years, and nineteen ninety-five Donoho and Johnstone propose wavelet threshold denoising method.Although the method is applied in a lot of places, and obtain good denoising effect, but still there is inevitable shortcoming in himself, particularly threshold value is chosen, want the wavelet coefficient after basis signal decomposes to weigh, the too small meeting of value causes signal section characteristic information and noise together by filtering, and distorted signals is serious; And the excessive denoising effect that makes of value is not obvious, signal to noise ratio (S/N ratio) can not effectively be improved.
In conjunction with the feature of two kinds of denoising methods, EMD and wavelet analysis are fused together and carry out signal processing, first utilize EMD to decompose low signal-to-noise ratio detection signal, time delay between adjacent signal peak point is defined as to time scale, and allow signal be decomposed into some IMF components of different scale, successively obtain the information of signal; Use again wavelet analysis to analyze one by one denoising to the each IMF component decompositing, preserve spike and sudden change part in signal, in narrower frequency band, the wavelet coefficient of useful signal is more outstanding, the more effective differentiation signal of energy and noise, choose more suitably threshold value and remove noise, thereby improve signal to noise ratio (S/N ratio).
Summary of the invention
The object of the invention is in order to strengthen detection signal signal to noise ratio (S/N ratio), improve the reliability of coarse grain material Ultrasonic Detection, a kind of coarse grain material supersonic detection method based on EMD associating wavelet threshold denoising is provided, fully advantages in detection signal is processed in conjunction with both, can carry out reasonable denoising to the detection signal of low signal-to-noise ratio, maximum stick signal characteristic information in removing noise, finally reaches coarse grain material is carried out to the object effectively detecting.
The present invention is achieved like this, technical scheme is: in conjunction with EMD, adaptive decomposition ability and the multi-scale wavelet of non-stationary nonlinear properties are differentiated to noise reduction capability more, the Low SNR signal that coarse grain material is detected carries out denoising, substantially improve detection signal signal to noise ratio (S/N ratio), realize the effective detection to coarse grain material; Be characterised in that concrete grammar is: in the coarse grain material of known defect, obtain ultrasound detection signal, because coarse grain material crystal grain is large, ultrasonic signal scattering and decay are serious, cause signal to noise ratio (S/N ratio) low; First detection signal is carried out to EMD processing, utilize the adaptive decomposition ability of EMD, decomposite from high to low intrinsic mode function (IMF) component of a series of narrow-bands by frequency, in each component, contain the characteristic information of original signal; Then utilize wavelet threshold denoising method to carry out denoising to each IMF component, because IMF component frequency band is narrow, small echo only need to carry out wavelet decomposition in little frequency band range, so the wavelet coefficient of useful signal is more outstanding, denoising effect is better, and the reconstruction signal of acquisition can obtain optimal estimation; Finally the each IMF after Wavelet Denoising Method is carried out to signal reconstruction, just can reach the object that retains as much as possible the useful feature information of detection signal on the basis of removing noise.
Advantage of the present invention is: fully advantages in detection signal is processed in conjunction with both, can carry out reasonable denoising to the detection signal of low signal-to-noise ratio, the more effective differentiation signal of energy and noise, choose more suitably threshold value and remove noise, significantly improve signal to noise ratio (S/N ratio), finally reach coarse grain material is carried out to the object effectively detecting.
Accompanying drawing explanation
Fig. 1 is oscillogram before and after emulation testing signal loading of the present invention.
Fig. 2 is the comparison diagram of three kinds of method results of the present invention and original signal global waveform.
Fig. 3 is coarse grain material test signal figure of the present invention.
Fig. 4 is EMD associating Wavelet Denoising Method result of the present invention and test signal comparison diagram.
In Fig. 1, upper partial graph is noiseless detection signal, and lower partial graph is for adding detection signal after noise, in Fig. 1 by left-to-right be beginning ripple, defect waves and end ripple.
Embodiment
Below in conjunction with accompanying drawing explanation, embodiments of the invention are described in further detail, but the present embodiment is not limited to the present invention, every employing analog structure of the present invention and similar variation thereof, all should list protection scope of the present invention in.
A kind of coarse grain material supersonic detection method based on EMD associating wavelet threshold denoising,, in conjunction with EMD, adaptive decomposition ability and the multi-scale wavelet of non-stationary nonlinear properties are differentiated to noise reduction capability more, the Low SNR signal that coarse grain material is detected carries out denoising, substantially improve detection signal signal to noise ratio (S/N ratio), realize the effective detection to coarse grain material; Be characterised in that concrete grammar is as follows: first coarse grain material detection signal of design of Simulation, add random noise to form by a noise-free signal, signal to noise ratio (S/N ratio) is 5dB, as shown in Figure 1, in figure, because signal to noise ratio (S/N ratio) is low, defect waves is almost flooded by noise, cannot be resolved out; Then detection signal is carried out to EMD decomposition, (if removed after the serious IMF component of front 3 noises according to EMD denoising method, overall signal to noise ratio (S/N ratio) is 11.13dB can to obtain 13 narrow-band IMF components and residual signals by adaptive decomposition from high to low of frequency; And directly carry out after wavelet threshold denoising, overall situation signal to noise ratio (S/N ratio) is 12.87dB), because these IMF component frequency bands are narrow, when they are carried out to wavelet threshold processing, the wavelet coefficient difference of noise and useful signal composition is obvious, choosing of threshold value is more convenient accurate, therefore removes the noise while in each IMF component, and what the characteristic information of signal can be more complete remains; Finally, by the IMF component reconstruction signal after denoising, calculating overall signal to noise ratio (S/N ratio) is 16.58dB, and signal to noise ratio (S/N ratio) improves a lot.
As shown in Figure 2, three kinds of methods are processed rear overall waveform and original noise-free signal comparison, and defect waves are being carried out to relatively can observe separately the integrated degree that various disposal routes retain for signal characteristic information.
According to this method characteristic, actual coarse grain material test block being detected to data processes, test block grain size is 2 grades, thickness is 80mm, having an aperture at buried depth 60mm is the flat-bottom hole of Φ 2mm, measured signal is as Fig. 3, begin in figure ripple and ground echo is more obvious, and flat-bottom hole reflection echo is flooded by noise.After this disposal route, by result and measured signal, more as shown in Figure 4, flat-bottom hole reflected signal " is shown one's talent " from very noisy signal, and reflection position is high-visible.
Claims (1)
1. the coarse grain material supersonic detection method based on EMD associating wavelet threshold denoising, technical scheme is: in conjunction with EMD, adaptive decomposition ability and the multi-scale wavelet of non-stationary nonlinear properties are differentiated to noise reduction capability more, the Low SNR signal that coarse grain material is detected carries out denoising, substantially improve detection signal signal to noise ratio (S/N ratio), realize the effective detection to coarse grain material; It is characterized in that concrete grammar is:
(1) detection signal is carried out to EMD processing, utilize the adaptive decomposition ability of EMD, decomposite from high to low intrinsic mode function (IMF) component of a series of narrow-bands by frequency, in each component, contain the characteristic information of original signal;
(2) then utilize wavelet threshold denoising method to carry out denoising to each IMF component, because IMF component frequency band is narrow, small echo only need to carry out wavelet decomposition in little frequency band range;
(3) the each IMF after Wavelet Denoising Method is carried out to signal reconstruction, just can reach the object that retains as much as possible the useful feature information of detection signal on the basis of removing noise.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410150224.0A CN103901115A (en) | 2014-04-15 | 2014-04-15 | Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410150224.0A CN103901115A (en) | 2014-04-15 | 2014-04-15 | Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103901115A true CN103901115A (en) | 2014-07-02 |
Family
ID=50992574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410150224.0A Pending CN103901115A (en) | 2014-04-15 | 2014-04-15 | Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103901115A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105954360A (en) * | 2016-06-03 | 2016-09-21 | 河北省电力建设调整试验所 | Ultrasonic testing method for coarse grain of 20 CrlMolVNbTiB high-temperature bolt |
| CN117368141A (en) * | 2023-12-07 | 2024-01-09 | 国检测试控股集团湖南华科科技有限公司 | Perchlorate wastewater concentration intelligent detection method based on artificial intelligence |
-
2014
- 2014-04-15 CN CN201410150224.0A patent/CN103901115A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105954360A (en) * | 2016-06-03 | 2016-09-21 | 河北省电力建设调整试验所 | Ultrasonic testing method for coarse grain of 20 CrlMolVNbTiB high-temperature bolt |
| CN117368141A (en) * | 2023-12-07 | 2024-01-09 | 国检测试控股集团湖南华科科技有限公司 | Perchlorate wastewater concentration intelligent detection method based on artificial intelligence |
| CN117368141B (en) * | 2023-12-07 | 2024-03-01 | 国检测试控股集团湖南华科科技有限公司 | Perchlorate wastewater concentration intelligent detection method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Transient signal analysis based on Levenberg–Marquardt method for fault feature extraction of rotating machines | |
| US7657403B2 (en) | Method for reducing digital data in an EMAT pig | |
| CN102636577B (en) | Method for extracting rim ultrasonic flaw detection signal on basis of wavelet transformation | |
| Li et al. | Extraction of oil debris signature using integral enhanced empirical mode decomposition and correlated reconstruction | |
| Shi et al. | Non-destructive testing of full-length bonded rock bolts based on HHT signal analysis | |
| US7779693B2 (en) | Method for nondestructive testing of pipes for surface flaws | |
| US10436832B2 (en) | Method of analyzing a cable, based on an auto-adaptive correlation, for the detection of soft defects | |
| JP5105384B2 (en) | Nondestructive inspection method and apparatus | |
| Qin et al. | Adaptively detecting the transient feature of faulty wind turbine planetary gearboxes by the improved kurtosis and iterative thresholding algorithm | |
| CN114062490B (en) | GAN-based rail weld Lamb wave modal decomposition and crack damage monitoring method | |
| Li et al. | Acoustic microscopy signal processing method for detecting near-surface defects in metal materials | |
| CN103792287A (en) | Large-area structural damage detection method based on Lamb wave | |
| CN107315048A (en) | A kind of steel member absolute stress recognition methods based on shear wave spectrum analysis | |
| Hao et al. | Detection method based on Kalman filter for high speed rail defect AE signal on wheel-rail rolling rig | |
| CN110412129A (en) | A kind of microscopic defect ultrasound detection signal processing method considering surface roughness | |
| Tang et al. | A method based on SVD for detecting the defect using the magnetostrictive guided wave technique | |
| Song et al. | A real time nondestructive crack detection system for the automotive stamping process | |
| Zárate et al. | Structural health monitoring of liquid-filled tanks: a Bayesian approach for location of acoustic emission sources | |
| Kehlenbach et al. | Identifying damage in plates by analyzing Lamb wave propagation characteristics | |
| JP2003066015A (en) | Signal processing method in acoustic emission method | |
| CN103901115A (en) | Ultrasonic coarse grain material detection method based on EMD (empirical mode decomposition) and wavelet threshold denoising | |
| Qi et al. | Study of submarine pipeline corrosion based on ultrasonic detection and wavelet analysis | |
| CN105929029A (en) | Noise processing method for SH guided wave nondestructive test technology | |
| Mao et al. | Noise suppression and flaw detection of ultrasonic signals via empirical mode decomposition | |
| Xu et al. | A novel method of bolt detection based on variational modal decomposition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140702 |
|
| WD01 | Invention patent application deemed withdrawn after publication |