CN109884192A - Sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction - Google Patents
Sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction Download PDFInfo
- Publication number
- CN109884192A CN109884192A CN201910155387.0A CN201910155387A CN109884192A CN 109884192 A CN109884192 A CN 109884192A CN 201910155387 A CN201910155387 A CN 201910155387A CN 109884192 A CN109884192 A CN 109884192A
- Authority
- CN
- China
- Prior art keywords
- signal
- echo
- weld seam
- atom
- guided wave
- 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.)
- Granted
Links
Landscapes
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The invention discloses a kind of sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction, the method for the present invention includes: to propose a new defect characteristic extracting method using the sparsity of echo-signal；Construct the echo-signal sparse representation model an of Parameter adjustable；During the realization of model, the Morlet small echo atom most like with pumping signal is used to construct complete atom dictionary for base, then solves base using division augmentation Lagrange contraction algorithm and track denoising model；According to the wave crest arrival time in reconstruction signal obtained by rarefaction representation, the accurate positionin of defective locations is realized；Simulation and actual welds flaw indication demonstrate the validity of this method.
Description
Technical field
The present invention relates to a kind of analysis process fields of signal, belong to non-destructive testing more particularly to a kind of strong background noise
The rarefaction representation detection method and defect inspection method of each ingredient, can be used for cut deal and large pressurized vessel etc. and lead in lower signal
The feature extraction and defects detection of wave echo-signal.
Background technique
Ultrasonic guided wave detection technology has been widely used for field of non destructive testing.Feature guided wave technology is applied to weld seam
Defects detection, when occurring local defect in weld seam, since guided wave can return to flaw echo wave packet when encountering defect.In wave packet
The echo-signal of each mode is contained, wherein carrying the structural information and defect information of weld seam.However, flaw echoes are logical
It often will receive seriously affecting for ambient noise, it is therefore desirable to develop advanced signal processing technology and fall vacant to eliminate noise and detection
Fall into signal.
There are three ways to commonly used to handle echo-signal, i.e. time-domain analysis, frequency-domain analysis, time frequency analysis.Time domain point
Analysis can accurately analyze the distribution of signal time domain energy but the influence vulnerable to mode conversion and dispersion phenomenon.Frequency-domain analysis is steady in processing
When signal, frequency domain energy distribution can be directly acquired, is distinguished with different modalities contained in frequency band.However, characteristics of weld seam guided wave
Signal always non-stationary signal, this only to be difficult accurately to extract defect characteristic with frequency-domain analysis.
Summary of the invention
To solve the above problems, the invention proposes a kind of sparsities using waveform to extract characteristics of weld seam guided wave echo
The new method of the defects of signal feature.The present invention can extract the defects of signal feature, can be used for thick in detection
The feature extraction and defects detection of the feature guided wave echo-signal of weld seam in plate.
To achieve the goals above, the present invention provides following technological means:
A kind of sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction, including method:
1) echo-signal for obtaining characteristics of weld seam guided wave establishes sparse representation model to detection signal；
2) the optimal Morlet wavelet basis atom most like with original signal, and then constitution step 1 are obtained using correlation filtering)
Excessively complete atom dictionary A needed for middle sparse representation model；
3) it is solved with base tracking Denoising Problems of the division augmentation Lagrange contraction algorithm to introducing, obtains step 1)
Sparse vector estimated value needed for middle sparse representation modelThe algorithm is by being updated iteration to sparse vector c, so that generation
Valence function J is minimized, and eventually finds one group of optimal solution；
4) by step 2) and step 3) A andThe rarefaction representation of detection signal is obtained, and calculates the peak value of echo wave packet
Arrival time extracts flaw echo, positions defect.
Further, the echo-signal for obtaining characteristics of weld seam guided wave, it includes following for establishing sparse representation model to detection signal
Step:
It is located at the stimulus sensor on weld seam top by function generator excitation, forms guided wave in weld seam, signal passes through
Weld seam to be measured is propagated, and sensor reception is received, then is sent into digital oscilloscope after preamplifier amplifies and is shown, is as detected
Signal y；
The relational model of detection signal y and original signal x and noise signal n is established, and with the combination of a linear element
Indicate unknown signaling:
X=Ac
Wherein, A indicates the matrix dictionary of N × M, is a matrix, and it is all vector, use is just thick that c, which indicates rarefaction representation coefficient,
Body surface shows.
Further, optimal Morlet wavelet basis atomic structre one mistake most like with original signal is obtained using correlation filtering
The process of complete atom dictionary A are as follows: define Morlet small echo atom ψγ(t), using correlation filtering method, matching is obtained and original
The sub- ψ of the most like base of signal (f, ζ, τ, t) obtains atom dictionary A using the time shift and transformation of ψ (f, ζ, τ, t).
Further, optimal Morlet wavelet basis atomic structre one mistake most like with original signal is obtained using correlation filtering
The detailed process of complete atom dictionary A are as follows:
Step S301: according to the bilateral fading characteristics of pumping signal and feature guided wave echo-signal, choosing has class with it
Like attenuation properties Morle small echo as the atom in atom dictionary A, Morlet small echo is defined as:
In formula, f ∈ R+For frequency of oscillation；For viscous damping ratio；τ ∈ R is time parameter.
Step S302: pass through the available wavelet basis ψ of correlation filteringγ(t) optimal when most like with original signal x (t)
Match parameter, related coefficient is bigger, indicates that matching is better, related coefficient CvExpression formula is
By maximizing related coefficient C corresponding to each time valuevIt calculates to obtain the small baud of most like Morlet
Levy parameterFor the C of given timevCorresponding maximum value is
By maximizing coefficient kγThe value of (τ) can obtain time parameter τ；
Step S303: obtaining the optimal value of the parameter of f, ζ, τ according to correlation filtering method, constructs most like with original signal
Morlet small echo basis function, and it is extended for the small echo atom dictionary A with different time shift parameters.
Further, division augmentation Lagrange contraction algorithm can solve Optimal solution problem by the method for iteratively faster:
Using the thought of separating variables, BPD (base tracking denoising) model is introduced by unconstrained optimization problem and is converted into constraint
Optimization problem；
By being updated iteration to sparse vector, so that cost function minimizes, one group of optimal solution is eventually found.
Further, BPD model is solved to division augmentation Lagrange contraction algorithm, specifically included:
By being updated iteration to sparse vector c, so that cost function J is minimized, one group of optimal solution is eventually found:
By thought separating variables, above formula can be converted to constrained optimization problem:
Wherein,G=I, " s.t. " is " subject in formula
The abbreviation of to " indicates the meaning of " satisfaction ", and λ is Lagrange multiplier, and formula indicates, under the premise of meeting Gc=v, so that
f1(c)+f2(v) the smallest vector c is the solution of the equation；
It takesB=0, H=(G-I), then above formula can be converted intoSolve the Shi Ke get
dk+1=dk-(Hzk+1-b)
Again by the above-mentioned equation of iterative solution, new equation is obtained:
dk+1=dk-(ck+1-vk+1-b)
Wherein, k is the number of iterations, and μ is specified penalty factor, is termination condition by the way that given the number of iterations is arranged,
Finally export optimal solutionThen actual signal x (t) approaches value and can be expressed as
Further, flaw echo is extracted, the detailed process of defect is positioned are as follows:
Reconstruction signal is obtained by rarefaction representation, and extracts initial spike arrival time t1, edge echo arrival time t2With lack
Fall into echo arrival time td, so that distance L of the defect apart from weld seam front end be calculatedd, position defect:
Wherein, LwFor fusion length, l is position of the receiving sensor apart from weld seam front end.
The present invention using sensor excitation characteristics of weld seam Guided waves weld seam and receives echo-signal, using echo-signal as
Detect signal；The detection signal is analyzed using detection method, analyzes echo-signal, solves sparse representation model,
Obtain sparse vector, reconstruction defect echo-signal；Flaw echo is identified in the echo-signal of reconstruct, obtains flaw echo
Moment positions defect.
As can be seen from the above technical solutions, the present invention implement provide a kind of sparsity for using echo-signal come
The detection technique for extracting defect characteristics has the advantage that
The echo-signal for analyzing characteristics of weld seam guided wave in the method for the present invention first constructs a sparse representation model.
This method constructs the dictionary of an expression sparse signal using Morlet wavelet basis function.Division augmentation Lagrange is shunk
Algorithm is applied to effectively solve BPD model, sparse vector is obtained, thus reconstruction defect echo-signal.Believe according to when reconstruction defect
Number occur position it is identical as real defect signal when, obtained wave crest arrival time, realize the accurate positionin of defective locations；Mould
It fits actual welds flaw indication and demonstrates the reliability and validity of this method.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the sparse representation model flow chart that ultrasonic echo feature disclosed by the embodiments of the present invention extracts；
Fig. 2 is the flow chart of construction atom dictionary disclosed by the embodiments of the present invention；
The processing result for the analog simulation research that Fig. 3 is signal-to-noise ratio disclosed by the embodiments of the present invention when being 5dB, wherein (a)
(b) it is noise signal figure for artificial echo signal waveforms, (c) is mixed waveform signal figure, is (d) the optimal of emulation signal
Atomic diagram is (e) the sparse vector figure of emulation signal, is (f) the reconstruction signal waveform diagram of emulation signal；
Fig. 4 is weld defect detection system figure disclosed by the embodiments of the present invention, wherein (a) is experimental provision schematic diagram, (b)
For pictorial diagram；
Fig. 5 is the processing result of signal obtained by weld defect detection system disclosed by the embodiments of the present invention, wherein (a) is to lead
Wave echo-signal waveform diagram, (b) the optimal atomic diagram of guided wave echo-signal are (c) the sparse vector figure of guided wave echo-signal,
It (d) is the reconstruction signal waveform diagram of guided wave echo-signal；
Fig. 6 is characteristics of weld seam guided waves propagation schematic diagram.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
It is an object of the present invention to provide a kind of defect inspection methods based on sparse model, for extracting and detecting echo-signal
The defects of signal.This method constructs the dictionary of an expression sparse signal using Morlet wavelet basis function.Division is increased
Wide Lagrange contraction algorithm is applied to effective solving model, obtains rarefaction representation coefficient vector, so that reconstruction defect echo is believed
Number.According to when reconstruction defect signal generation position is identical as real defect signal, obtained wave crest arrival time, realize scarce
Fall into the accurate positionin of position
As shown in Figure 1, the present invention includes the following steps:
Step S101: the echo-signal of characteristics of weld seam guided wave is obtained；
Wherein, it is located at the stimulus sensor on weld seam top when experiment by function generator excitation, is formed and led in weld seam
Wave, signal are propagated by weld seam to be measured, are received sensor reception, then are sent into digital oscilloscope after preamplifier amplifies and are shown
Show, as detection signal y.
Step S102: sparse representation model is established to the detection signal；
Establish the relational model of detection signal y and original signal x and noise signal n:
Y=x+n (1)
Unknown signaling is indicated with the combination of a linear element:
X=Ac (2)
Wherein, A indicates the matrix dictionary of N × M, is a matrix, and y is detection signal, and x is actual signal, and n indicates noise
Ingredient, c indicate rarefaction representation coefficient, are all vector, therefore indicated with positive runic.
Step S103: optimal Morlet wavelet basis atomic structre one most like with original signal is obtained using correlation filtering
Cross complete atom dictionary.
Wherein step S103 includes the selection of base and the acquisition of parameters again；
Step S301: according to the bilateral fading characteristics of pumping signal and feature guided wave echo-signal, choosing has class with it
Like attenuation properties Morle small echo as the atom in atom dictionary A.Morlet small echo is defined as:
In formula, f ∈ R+For frequency of oscillation；For viscous damping ratio；τ ∈ R is time parameter.
Step S302: pass through the available wavelet basis ψ of correlation filteringγ(t) optimal when most like with original signal x (t)
Match parameter.Related coefficient is bigger, indicates that matching is better, and related coefficient expression formula is
By maximizing related coefficient C corresponding to each time valuevIt can calculate to obtain most like Morlet small
Wave characteristic parameterFor the C of given timevCorresponding maximum value is
By maximizing coefficient kγThe value of (τ) can obtain time parameter τ.
Step S303: obtaining the optimal value of the parameter of f, ζ, τ according to correlation filtering method, constructs most like with original signal
Morlet small echo basis function, and it is extended for the small echo atom dictionary A with different time shift parameters.
Step S104: model is solved with division augmentation Lagrange contraction algorithm, is solvedThe algorithm by pair
Sparse vector c is updated iteration, so that cost function J is minimized, eventually finds one group of optimal solution.
Can be constrained optimization problem by formula (8) by thought separating variables:
Wherein,G=I, " s.t. " is in formula (7)
The abbreviation of " subject to " indicates the meaning of " satisfaction ", and λ is Lagrange multiplier.Formula (7) indicates, is meeting Gc=v's
Under the premise of, so that f1(c)+f2(v) the smallest vector c is the solution of the equation.
It takesB=0, H=(G-I), then formula (7) can be converted intoSolve the Shi Ke get
dk+1=dk-(Hzk+1-b) (9)
Again by the above-mentioned equation of iterative solution, new equation is obtained:
dk+1=dk-(ck+1-vk+1-b) (12)
Wherein, k is the number of iterations, and μ is specified penalty factor.It is termination condition by the way that given the number of iterations is arranged,
Finally export optimal solutionThen actual signal x (t) approaches value and can be expressed asvk、dkFor the assumption value generation of variable
Number, I is unit matrix；
Step S105: calculating the peak value arrival time of echo wave packet, extracts flaw echo, positions defect.
Initial spike arrival time t is obtained by reconstruction signal1, edge echo arrival time t2With flaw echo arrival time
td.To which distance L of the defect apart from weld seam front end be calculatedd, position defect.
Wherein, LwFor fusion length, l is position of the receiving sensor apart from weld seam front end
A kind of rarefaction representation for characteristics of weld seam guided wave flaw echoes feature extraction is proposed in the method for the present invention
Model.First construct echo-signal sparse representation model, wherein use the Morlet small echo atom most like with pumping signal for
Base constructed complete atom dictionary, then using division augmentation Lagrange contraction algorithm solving model, obtained reconstruction signal.Root
According to when reconstruction defect signal generation position is identical as real defect signal, obtained wave crest arrival time, defective bit is realized
The accurate positionin set.
In order to verify the validity of proposed method, the analog study of noise-containing emulation signal is elaborated
It states:
Pumping signal is the HANNING window modulated sinusoid signal in 5 periods, can be indicated are as follows:
It is constructed by the simulation of ABAQUS finite element software close to actual welding line structure and defect type combination pumping signal
Emulation signal, and white Gaussian noise n (t) is added wherein, obtain emulation testing signal y (t).
Signal-to-noise ratio can be used to measure noise level, is defined as:
Wherein, PsIt is the energy of signal, is represented by
PnIt is the energy of noise signal, is represented by
(a)-(c) in Fig. 3 is respectively the signal y of the echo-signal x (t) simulated, noise signal n (t), mixed noise
(t) waveform.Apply the inventive method to the analysis of analog signal and the defects detection of signal.The sparse of signal is constructed first
Indicate model, Morlet wavelet basis dictionary then is constructed to indicate original signal using correlation filtering, and optimal atom is represented by
Parameter is ψ (119700,0.18, τ, t).And sparse vector is solved with the division augmentation Lagrange contraction algorithm of 50 iteration,
Finally obtain reconstruction signal.λ=11, μ=3 are set herein.Fig. 3 (d)-(f) is respectively the most matched Morlet of original signal
The echo-signal that wavelet basis atom, sparse vector and reconstruct obtain.The generation moment that therefrom can intuitively obtain flaw echo is
td=029ms is, consistent with the flaw echo generation moment in analog signal, realizes defect characteristic and extracts and position.
Numerical procedure for a better understanding of the present invention, it is that two block sizes are that Fig. 4, which gives experimental subjects,
The 245# steel plate butt weld of 800mmx800mmx10mm is that the analysis that the data that experimental subjects obtains carry out is example, wherein welding
Sew on, be left it is high be respectively 3.6mm and 4mm away from weld seam front end 625mm at provided with the artificial mould of a diameter 3mm depth 8mm
Quasi- hole corrosion default tells about the application of sparse representation model method in the signal in detail:
Embodiment: the data obtained to experiment are analyzed
When experiment, 5 periods that centre frequency is 200~250kHz are generated as illustrated by signal generator (DG4062)
Hanning window modulation waveform, excitation are located at the stimulus sensor on weld seam top, SH1 mode are formed in weld seam, signal is by be measured
Weld seam is propagated, and is received sensor reception, then be sent into digital oscilloscope after the preamplifier amplification that gain is 40dB
(DS2102A) observation display.
It is shown in Fig. 5 (a) with flaw echo, it is clear that flaw echo cannot be identified from Fig. 5 (a).
Echo-signal is applied the inventive method to, process is identical as the analysis method that above-mentioned analogue simulation is studied.Herein set λ=
11, μ=3, the number of iterations of division augmentation Lagrange contraction algorithm is 50.Analyze success shown in result such as Fig. 5 (b)-(d)
Obtain flaw echoes.
Fig. 6 is characterized propagation schematic diagram of the guided wave in weld seam, obtains t by Fig. 5 (d)1=0.12ms, td=0.48ms,
t2=0.61mS.Known parameters are Lw=800mm, l=100mm, can be calculated defective locations is Ld=615mm, the phase with true value
It is 1.6% to error.It is almost the same with real defect position.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " illustrative examples ",
The description of " example ", " specific example " or " some examples " etc. means specific features described in conjunction with this embodiment or example, knot
Structure, material or feature are included at least one embodiment or example of the invention.In the present specification, to above-mentioned term
Schematic representation may not refer to the same embodiment or example.Moreover, specific features, structure, material or the spy of description
Point can be combined in any suitable manner in any one or more of the embodiments or examples.
Although an embodiment of the present invention has been shown and described, it will be understood by those skilled in the art that: not
A variety of change, modification, replacement and modification can be carried out to these embodiments in the case where being detached from the principle of the present invention and objective, this
The range of invention is defined by the claims and their equivalents.
Claims (7)
1. a kind of sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction characterized by comprising
1) echo-signal for obtaining characteristics of weld seam guided wave establishes sparse representation model to detection signal；
2) the optimal Morlet wavelet basis atom most like with original signal, and then constitution step 1 are obtained using correlation filtering) in it is dilute
Dredge atom dictionary A excessively complete needed for indicating model；
3) it is solved, is obtained dilute in step 1) with base tracking Denoising Problems of the division augmentation Lagrange contraction algorithm to introducing
Dredge sparse vector estimated value needed for indicating modelThe algorithm is by being updated iteration to sparse vector c, so that cost letter
Number J is minimized, and eventually finds one group of optimal solution；
4) by step 2) and step 3) A andThe rarefaction representation of detection signal is obtained, and the peak value for calculating echo wave packet reaches
Time extracts flaw echo, positions defect.
2. representation method according to claim 1, which is characterized in that the echo-signal for obtaining characteristics of weld seam guided wave, to inspection
Survey signal establish sparse representation model the following steps are included:
It is located at the stimulus sensor on weld seam top by function generator excitation, forms guided wave in weld seam, signal is by be measured
Weld seam is propagated, and sensor reception is received, then is sent into digital oscilloscope after preamplifier amplifies and is shown, as detection signal
y；
The relational model of detection signal y and original signal x and noise signal n is established, and is indicated with the combination of a linear element
Unknown signaling:
X=Ac
Wherein, A indicates the matrix dictionary of N × M, is a matrix, and it is all vector, with positive runic table that c, which indicates rarefaction representation coefficient,
Show.
3. representation method according to claim 1, which is characterized in that constructed the process of complete atom dictionary A are as follows: fixed
Adopted Morlet small echo atom ψγ(t), using correlation filtering method, matching obtain the sub- ψ of the base most like with original signal (f, ζ, τ,
T), atom dictionary A is obtained using the time shift and transformation of ψ (f, ζ, τ, t).
4. representation method according to claim 3, which is characterized in that constructed the detailed process of complete atom dictionary A
Are as follows:
Step S301: according to the bilateral fading characteristics of pumping signal and feature guided wave echo-signal, choosing has similar decline with it
Subtract the Morle small echo of property as the atom in atom dictionary A, Morlet small echo is defined as:
In formula, f ∈ R+For frequency of oscillation；For viscous damping ratio；τ ∈ R is time parameter.
Step S302: pass through the available wavelet basis ψ of correlation filteringγ(t) Optimum Matching with original signal x (t) when most like
Parameter, related coefficient is bigger, indicates that matching is better, related coefficient CvExpression formula is
By maximizing related coefficient C corresponding to each time valuevIt calculates to obtain most like Morlet wavelet character parameterFor the C of given timevCorresponding maximum value is
By maximizing coefficient kγThe value of (τ) can obtain time parameter τ；
Step S303: obtaining the optimal value of the parameter of f, ζ, τ according to correlation filtering method, constructs most like with original signal
Morlet small echo basis function, and it is extended for the small echo atom dictionary A with different time shift parameters.
5. representation method according to claim 1, which is characterized in that division augmentation Lagrange contraction algorithm passes through quick
The method of iteration solves Optimal solution problem:
Using the thought of separating variables, base tracking denoising model is introduced by unconstrained optimization problem and is converted into constrained optimization problem；
By being updated iteration to sparse vector, so that cost function minimizes, one group of optimal solution is eventually found.
6. representation method according to claim 1, which is characterized in that solve base with division augmentation Lagrange contraction algorithm
Denoising model is tracked, is specifically included:
Iteration is updated to sparse vector c by dividing augmentation Lagrange contraction algorithm, so that cost function J is minimized,
Eventually find one group of optimal solution:
By thought separating variables, formula can be converted to constrained optimization problem:
Wherein,f2(v)=λ | | c | |1, G=I, the meaning of " s.t. " expression " satisfaction " in formula
Think, λ is Lagrange multiplier, and formula indicates under the premise of meeting Gc=v, so that f1(c)+f2(V) the smallest vector c is to be somebody's turn to do
Non trivial solution；
Take E (z)=f1(c)+f2(v),B=0, H=(G-I), then above formula can be converted intoSolve the Shi Ke get
dk+1=dk-(Hzk+1-b)
Again by the above-mentioned equation of iterative solution, new equation is obtained:
dk+1=dk-(ck+1-vk+1-b)
Wherein, k is the number of iterations, and μ is specified penalty factor, is termination condition by the way that given the number of iterations is arranged, finally
Export optimal solutionThen actual signal x (t) approaches value and can be expressed as
7. representation method according to claim 1, it is characterised in that: calculate the peak value arrival time of echo wave packet, extract
Flaw echo positions the detailed process of defect are as follows:
Initial spike arrival time t is obtained by reconstruction signal1, edge echo arrival time t2With flaw echo arrival time td, from
And distance L of the defect apart from weld seam front end is calculatedd, position defect:
Wherein, LwFor fusion length, l is position of the receiving sensor apart from weld seam front end.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910155387.0A CN109884192B (en) | 2019-02-28 | 2019-02-28 | Sparse representation method for extracting characteristics of defect echo signals of weld characteristic guided waves |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910155387.0A CN109884192B (en) | 2019-02-28 | 2019-02-28 | Sparse representation method for extracting characteristics of defect echo signals of weld characteristic guided waves |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109884192A true CN109884192A (en) | 2019-06-14 |
| CN109884192B CN109884192B (en) | 2022-04-26 |
Family
ID=66930204
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910155387.0A Active CN109884192B (en) | 2019-02-28 | 2019-02-28 | Sparse representation method for extracting characteristics of defect echo signals of weld characteristic guided waves |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109884192B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111272274A (en) * | 2020-02-22 | 2020-06-12 | 西北工业大学 | Closed space low-frequency sound field reproduction method based on microphone random sampling |
| CN112149508A (en) * | 2020-08-25 | 2020-12-29 | 江苏大学 | Method for extracting sparse representation of weld defect echo signal features based on shear wave characteristic fusion |
| CN112362343A (en) * | 2020-10-28 | 2021-02-12 | 华南理工大学 | Distributed fault feature extraction method for gearbox under variable rotating speed based on frequency modulation dictionary |
| CN112515637A (en) * | 2020-12-02 | 2021-03-19 | 山东省人工智能研究院 | Electrocardiosignal noise reduction method based on group sparsity characteristic |
| CN113514548A (en) * | 2021-07-09 | 2021-10-19 | 厦门大学 | Weld defect detection method based on characteristic guided waves |
| CN113884567A (en) * | 2021-06-29 | 2022-01-04 | 北京交通大学 | Steel rail weld damage detection method and device based on ultrasonic Lamb waves |
| CN116087345A (en) * | 2023-04-06 | 2023-05-09 | 广东电网有限责任公司揭阳供电局 | Method, device and medium for calculating axial defect length of composite insulator |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070173699A1 (en) * | 2006-01-21 | 2007-07-26 | Honeywell International Inc. | Method and system for user sensitive pacing during rapid serial visual presentation |
| CN101109732A (en) * | 2007-08-08 | 2008-01-23 | 哈尔滨工业大学 | Ultrasound nondestructive detecting echo signal classificating method based on vague plane characteristic |
| CN103792000A (en) * | 2014-02-20 | 2014-05-14 | 苏州大学 | Method and device for detecting transient components in signal based on sparse representation |
| CN106370734A (en) * | 2016-08-21 | 2017-02-01 | 安徽科技学院 | Method for extracting end surface reflection signal in ultrasonic guided-wave detection of pipeline |
| CN108376233A (en) * | 2018-01-11 | 2018-08-07 | 江苏大学 | A kind of the separation sparse representation method and inaction interval detection method of fault detect |
-
2019
- 2019-02-28 CN CN201910155387.0A patent/CN109884192B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070173699A1 (en) * | 2006-01-21 | 2007-07-26 | Honeywell International Inc. | Method and system for user sensitive pacing during rapid serial visual presentation |
| CN101109732A (en) * | 2007-08-08 | 2008-01-23 | 哈尔滨工业大学 | Ultrasound nondestructive detecting echo signal classificating method based on vague plane characteristic |
| CN103792000A (en) * | 2014-02-20 | 2014-05-14 | 苏州大学 | Method and device for detecting transient components in signal based on sparse representation |
| CN106370734A (en) * | 2016-08-21 | 2017-02-01 | 安徽科技学院 | Method for extracting end surface reflection signal in ultrasonic guided-wave detection of pipeline |
| CN108376233A (en) * | 2018-01-11 | 2018-08-07 | 江苏大学 | A kind of the separation sparse representation method and inaction interval detection method of fault detect |
Non-Patent Citations (2)
| Title |
|---|
| 吴彪: "超声无损检测缺陷定位与稀疏信号重构方法", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 樊薇: "信号瞬态成分稀疏表示方法及其机械故障特征提取应用研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111272274A (en) * | 2020-02-22 | 2020-06-12 | 西北工业大学 | Closed space low-frequency sound field reproduction method based on microphone random sampling |
| CN112149508A (en) * | 2020-08-25 | 2020-12-29 | 江苏大学 | Method for extracting sparse representation of weld defect echo signal features based on shear wave characteristic fusion |
| CN112149508B (en) * | 2020-08-25 | 2024-05-17 | 江苏大学 | Method for extracting sparse representation of echo signal characteristics of weld defects based on shear wave characteristic fusion |
| CN112362343A (en) * | 2020-10-28 | 2021-02-12 | 华南理工大学 | Distributed fault feature extraction method for gearbox under variable rotating speed based on frequency modulation dictionary |
| CN112515637A (en) * | 2020-12-02 | 2021-03-19 | 山东省人工智能研究院 | Electrocardiosignal noise reduction method based on group sparsity characteristic |
| CN113884567A (en) * | 2021-06-29 | 2022-01-04 | 北京交通大学 | Steel rail weld damage detection method and device based on ultrasonic Lamb waves |
| CN113884567B (en) * | 2021-06-29 | 2023-12-08 | 北京交通大学 | Rail weld damage detection method and device based on ultrasonic Lamb wave |
| CN113514548A (en) * | 2021-07-09 | 2021-10-19 | 厦门大学 | Weld defect detection method based on characteristic guided waves |
| CN116087345A (en) * | 2023-04-06 | 2023-05-09 | 广东电网有限责任公司揭阳供电局 | Method, device and medium for calculating axial defect length of composite insulator |
| CN116087345B (en) * | 2023-04-06 | 2023-06-13 | 广东电网有限责任公司揭阳供电局 | Method, device and medium for calculating axial defect length of composite insulator |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109884192B (en) | 2022-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109884192A (en) | Sparse representation method for characteristics of weld seam guided wave flaw echoes feature extraction | |
| CN106770664B (en) | A method of edge defect detection is improved based on total focus imaging algorithm | |
| CN109782274B (en) | Water damage identification method based on time-frequency statistical characteristics of ground penetrating radar signals | |
| CN106226407B (en) | A kind of online preprocess method of ultrasound echo signal based on singular spectrum analysis | |
| Drai et al. | Elaboration of some signal processing algorithms in ultrasonic techniques: application to materials NDT | |
| CN106841403A (en) | A kind of acoustics glass defect detection method based on neutral net | |
| CN103245726B (en) | Method for detecting material hydrogen damage through ultrasonic surface waves | |
| CN106864477B (en) | A kind of rail flaw ultrasonic detection intelligent detecting method based on time-frequency do-nothing function | |
| CN105004792A (en) | Nonlinear ultrasonic phased array imaging method for micro-crack detection | |
| CN114062490B (en) | GAN-based rail weld Lamb wave modal decomposition and crack damage monitoring method | |
| CN109085244A (en) | A kind of non-linear Lamb wave structure fatigue damage chromatography imaging method based on piezoelectric-array | |
| Li et al. | Acoustic microscopy signal processing method for detecting near-surface defects in metal materials | |
| CN105675126B (en) | A kind of new method for stablizing sound field acoustic pressure for detecting multifrequency multi-source complexity | |
| CN204758542U (en) | Detection apparatus for metallic structure's crackle | |
| Tang et al. | A method based on SVD for detecting the defect using the magnetostrictive guided wave technique | |
| CN116106690A (en) | Method for improving cable defect time domain recovery waveform characteristics based on low-frequency stitching and application | |
| Avanesians et al. | Wave separation: Application for arrival time detection in ultrasonic signals | |
| Fan et al. | Feature extraction of echo signal of weld defect guided waves based on sparse representation | |
| CN106054028B (en) | A kind of cable fault automatic range method based on temporal signatures and wavelet analysis | |
| He et al. | Quantitative detection of surface defect using laser-generated Rayleigh wave with broadband local wavenumber estimation | |
| CN205844271U (en) | A kind of detecting system based on difference frequency non-linear ultrasonic detection sheet metal micro-crack | |
| CN101726545B (en) | Method for extracting time-frequency amplitude characteristic and time-frequency phase characteristic of ultrasonic signals on dissimilar material diffusion welding interface | |
| Mao et al. | Noise suppression and flaw detection of ultrasonic signals via empirical mode decomposition | |
| Kažys et al. | Ultrasonic detection of defects in strongly attenuating structures using the Hilbert–Huang transform | |
| Yang et al. | Feature extraction of ultrasonic guided wave weld detection based on group sparse wavelet transform with tunable Q-factor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |