CN113884581A - Noise reduction method of ultrasonic detection signal based on additive defect - Google Patents

Noise reduction method of ultrasonic detection signal based on additive defect Download PDF

Info

Publication number
CN113884581A
CN113884581A CN202010635245.7A CN202010635245A CN113884581A CN 113884581 A CN113884581 A CN 113884581A CN 202010635245 A CN202010635245 A CN 202010635245A CN 113884581 A CN113884581 A CN 113884581A
Authority
CN
China
Prior art keywords
signal
defect
ultrasonic
ultrasonic detection
noise reduction
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
Application number
CN202010635245.7A
Other languages
Chinese (zh)
Other versions
CN113884581B (en
Inventor
赵吉宾
何振丰
赵宇辉
孙长进
王志国
李论
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenyang Institute of Automation of CAS
Original Assignee
Shenyang Institute of Automation of CAS
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenyang Institute of Automation of CAS filed Critical Shenyang Institute of Automation of CAS
Priority to CN202010635245.7A priority Critical patent/CN113884581B/en
Publication of CN113884581A publication Critical patent/CN113884581A/en
Application granted granted Critical
Publication of CN113884581B publication Critical patent/CN113884581B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
    • G01N29/4472—Mathematical theories or simulation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P10/00—Technologies related to metal processing
    • Y02P10/25—Process efficiency

Landscapes

  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)

Abstract

The invention discloses a noise reduction method of an ultrasonic detection signal based on additive defects, and belongs to the technical field of additive manufacturing. And denoising the ultrasonic detection signal of the additive defect by adopting a discrete wavelet transform algorithm, and obtaining the de-denoised detection signal through signal sampling, discrete wavelet decomposition, signal processing and discrete wavelet reconstruction. A threshold function which can be dynamically adjusted based on hard and soft threshold methods is adopted in the signal processing process. The method comprises the steps of prefabricating a laser additive material sample piece containing flat-bottom holes, air holes and crack additive material defects for detecting and extracting ultrasonic detection signals, evaluating the performance of the noise reduction algorithm, and finding that the noise reduction algorithm is superior to a traditional minimax soft threshold method by comparing noise reduction signal waveforms obtained by different noise reduction methods and the signal-to-noise ratio and root-mean-square error of the signals.

Description

Noise reduction method of ultrasonic detection signal based on additive defect
Technical Field
The invention relates to the technical field of additive manufacturing ultrasonic nondestructive testing, in particular to a noise reduction method of an ultrasonic detection signal based on additive defects.
Background
The metal additive manufacturing technology can directly form any complex part, saves the design and manufacturing cost of the traditional die clamp, shortens the processing period and has great advantages for manufacturing aerospace parts. However, due to the fact that defects such as poor fusion, air holes and cracks may be generated in the additive manufacturing technology due to the special forming process, and the internal defects have different properties from those of the traditional cast-forged piece, the detection standard of the cast-forged piece cannot be directly applied to the detection process, and a nondestructive detection technology and scheme suitable for the additive manufacturing field need to be developed urgently. The invention provides a signal noise reduction processing method aiming at analyzing an ultrasonic detection signal of an additive defect.
The maximum resolution of the ultrasonic detection theory is half of the wavelength of the ultrasonic wave (namely lambda/2), and when the defect with a small reflection interface needs to be detected, an ultrasonic probe with a large central frequency needs to be selected. The higher the ultrasonic frequency is, the weaker the penetration ability is, and when the defect reflected sound wave is weaker, a stronger gain needs to be given to the received signal, the electronic white noise is further increased, the defect sound wave is mixed in random noise and is difficult to distinguish, and whether a useful signal, namely a defect reflected wave, exists is more difficult to judge. It is therefore necessary to reduce the noise of the reflected wave signal.
Disclosure of Invention
Aiming at the detection requirements of the existing additive manufacturing technology, the invention provides a noise reduction method of an ultrasonic detection signal based on additive defects, which improves the signal-to-noise ratio of a defect echo under high gain by carrying out noise reduction processing on the ultrasonic detection signal of the additive defects so as to improve the identification and detection capability of small-size defects.
In order to achieve the purpose, the invention adopts the technical scheme that:
a noise reduction method of an ultrasonic detection signal based on an additive defect comprises the following steps: offsetting the original signal with the defect and the signal without the defect and removing the reflected wave signals on the upper surface and the lower surface to obtain the ultrasonic reflected wave signals of the workpiece to be measured; and performing discrete wavelet denoising treatment on the ultrasonic reflected wave signal to improve the signal to noise ratio, so as to obtain an ultrasonic detection signal of the defect and identify the material increase defect.
The method is carried out by utilizing an ultrasonic detection system, wherein the ultrasonic detection system comprises phased array integrated instrument equipment, an upper computer and an ultrasonic probe, the upper computer outputs signals to enable the phased array integrated instrument equipment to control the ultrasonic probe to transmit and collect original echo signals, and the upper computer processes and identifies defects. And the upper computer conducts signal derivation and processing through Focus PC data acquisition and analysis software and Matlab software.
The method for denoising an ultrasonic detection signal based on an additive defect according to claim 1, wherein the discrete wavelet denoising process comprises:
discrete wavelet decomposition: finding the best approximation f of the signal fjThen, down-sampling decomposition is carried out step by step until all wavelet components are obtained;
signal processing: eliminating partial frequency in wavelet component by threshold value quantization method to achieve the purpose of signal noise reduction;
and (3) discrete wavelet reconstruction: and performing up-sampling reconstruction on the wavelet component after signal processing to obtain a function to obtain an output signal of discrete wavelet de-noising processing, wherein the up-sampling reconstruction is an inverse process of discrete wavelet decomposition.
The discrete wavelet decomposition includes:
a. initialization: the purpose is to find the best approximation f of the ultrasonic reflected wave signal fjDue to fjFor signal f in approximate space VjOrthogonal projection P ofjf, when the scale function phi is tightly supported and the number of decomposition layers j is sufficiently large, there are:
Figure BDA0002568574100000031
Figure BDA0002568574100000032
where j is 1,2,3,4,5,6, m is an adjustment coefficient, m may be 1,2,3,4,5,6 … …, and the larger j is, the higher the approximation degree is, and f is, the larger f is, the higher f is, the lower the adjustment coefficient is, and the lower the adjustment coefficient is, the higher the adjustment coefficient is, and the lower the adjustment coefficient is, and the higher the adjustment coefficient isjThe accuracy of (f) depends on the smoothness of j and k, and f;
b. iteration: is to best approximate the signal fjCarrying out a downsampling decomposition process step by step;
for the best approximation signal fjDecomposed into the approximate part f of the next stage by a low-pass filterj-1The response of the low-pass filter is g [ n ]](ii) a Decomposition into wavelet portions w of the next stage by a high-pass filterj-1The response of the high-pass filter is h [ n ]]；
Figure BDA0002568574100000033
Figure BDA0002568574100000034
c. And (4) terminating: to fj-1The above decomposition process is repeated until all sample points are exhausted when j is 0.
The signal processing includes:
adopting the following threshold function to filter out electronic white noise and retain potential defect echo:
Figure BDA0002568574100000035
wherein, λ is a threshold value,
Figure BDA0002568574100000041
for noise-reduced signals, wjFor signals before noise reduction processing, m and n are adjustable coefficients, and the assignment of m and n is changed to enable the threshold function to approach to one direction between the soft threshold function and the hard threshold function.
The discrete wavelet signal reconstruction is as follows: obtained after processing the signalAll j wavelet signals w0，w1，……，wj-2，wj-1Using high-pass filtering, for an approximation signal f0The upsampling is performed using low pass filtering, resulting in the signal f.
The method can be used for detecting laser additive workpieces with flat bottom holes, air holes and crack defects.
Further comprising: and acquiring the waveform of the ultrasonic detection signal of the defect, calculating the signal-to-noise ratio and the root-mean-square error index, and comparing the signal-to-noise ratio and the root-mean-square error index with a minimax soft threshold method to verify that the method is effective.
After the technical scheme is adopted, the invention has the beneficial effects that:
1. the noise reduction method can obviously improve the signal-to-noise ratio of the signal under the high-gain condition and improve the detection rate of small-size defects.
2. The method has good adaptability to the noise reduction of the ultrasonic detection signal of the material increase defect.
Drawings
FIG. 1 is a schematic diagram of an ultrasonic inspection system;
FIG. 2 is a process block diagram;
FIG. 3 is a signal decomposition diagram;
FIG. 4 is a schematic diagram of signal reconstruction;
FIG. 5 is a waveform of an original signal extracted by ultrasonic testing
FIG. 6 shows the result of noise reduction processing of the ultrasonic reflected wave signal of defect 1;
FIG. 7 shows the result of noise reduction processing on the ultrasonic reflected wave signal of defect 2;
FIG. 8 shows the result of noise reduction processing of the ultrasonic echo signal of defect 3;
FIG. 9 shows the result of noise reduction processing on the ultrasonic reflected wave signal of defect 4;
FIG. 10 shows the result of noise reduction processing on the ultrasonic echo signal of the defect 5;
FIG. 11 shows the result of noise reduction processing on the ultrasonic reflected wave signal of defect 6;
FIG. 12 shows the result of noise reduction processing of the ultrasonic echo signal of the defect 7;
FIG. 13 is a comparison graph of signal-to-noise ratio for a noise reduction algorithm;
FIG. 14 is a root mean square error comparison plot of a noise reduction algorithm.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in detail with reference to the accompanying drawings and examples.
A noise reduction method of an ultrasonic detection signal based on an additive defect comprises the following steps:
step 1: testing and extracting an original ultrasonic detection signal;
step 2: discrete wavelet denoising processing is carried out on the sampled signals;
and step 3: and (5) evaluating the performance of the algorithm.
This will be described in detail below.
S1: original ultrasonic detection signal testing and extraction
The ultrasonic detection system used by the invention comprises an Olympus phased array integrated instrument device FPX-1664PR, Focus PC data acquisition and analysis software and a 15MHz probe V113-RM, and is shown in figure 1. The couplant used in the detection process is engine oil, the detection surface of the workpiece to be detected needs to be processed in advance, impurities are removed, and the surface roughness Ra is guaranteed to be less than or equal to 6.3 microns. The detection data is collected and exported by Focus PC data collection and analysis software.
If the workpiece to be measured is used as a system, the incident ultrasonic wave u (t) is the input signal of the system, γ (t) is the response function of the system, and the original output signal s (t) with defects of the system can be obtained by the convolution integral method:
Figure BDA0002568574100000061
wherein f (t) is the system output signal, i.e. the reflected wave signal containing the structure noise received by the ultrasonic probe, u (t) is the system input signal, gamma (t) is the response function of the system, RiIs a reflection coefficient, alpha is an ultrasonic attenuation coefficient, B is an ultrasonic probe bandwidth, omega is an angular frequency, tau is a reflected wave transit time, and t is a time variable.
Random noise of ultrasound reflected wavesThe noise comprises two parts of structural noise and electronic white noise. For structural noise, R in the above formulaiAnd τiIs a random variable. When the number of these random variables is large enough, the results fit a normal distribution. After the structure noise is modulated, the amplitude of the structure noise is almost unchanged, and the structure noise can be approximately regarded as electronic white noise. From the above equation, the echo signal of the ultrasonic detection is a non-stationary signal with respect to time t.
The discrete wavelet transform can perform time domain and frequency domain analysis, and is suitable for analyzing ultrasonic echo signals. This process is generally divided into 4 steps: sampling, decomposing, signal processing and reconstructing. The method structure is shown in fig. 2.
If the signal is continuous, the sampling frequency should be 2 times the original signal frequency according to Nyquist's law to ensure that the signal is not distorted. For the research, the original ultrasonic signals are derived by FOCUS PC data acquisition and analysis software, and discrete defect reflected wave signal data can be directly output.
S2: discrete wavelet de-noising processing of sampled signals
The discrete wavelet denoising processing of the sampling signal is divided into three aspects of signal decomposition, signal processing and reconstruction.
(1) Discrete wavelet transform decomposition algorithm
This stage splits the signal into components of several frequencies according to the following algorithm. The decomposition algorithm is divided into 3 steps: initialization, iteration and termination. The purpose of the initialization is to find the best approximation f of the ultrasonic reflected wave signal fjFrom an energy point of view, fjFor signal f in approximate space VjOrthogonal projection P ofjf, when the scale function φ is tightly supported and the number of decomposition layers j is sufficiently large:
Figure BDA0002568574100000071
Figure BDA0002568574100000072
wherein j is 1,2,3,4,5,6, m is an adjustment coefficient, m can be 1,2,3,4,5,6 … …, the larger the j is, the higher the approximation degree is, f isjThe accuracy of (f) depends on the smoothness of j and k, and f.
The iteration is tojAnd (5) a step-by-step decomposition process. For a certain j, fjCan be passed through a low pass filter (scaling filter, with response g n]) Decomposed into an approximation part f of the next stagej-1(fjAnd g [ n ]]Convolved and downsampled); passing through a high-pass filter (wavelet filter, with response h n]) Wavelet portion w decomposed into next levelj-1(fjAnd g [ n ]]Convolution and downsampling of):
Figure BDA0002568574100000073
Figure BDA0002568574100000074
wherein n is 0,1, 2,3,4,5, 6.
After passing through two filters of high and low pass and down sampling of step 2 respectively, fj-1And wj-1Signal bandwidths of all fjHalf, which means that both will lose half of the data samples. To fj-1The decomposition process is repeated until all sample points are exhausted when j is 0, and fig. 3 is a schematic diagram of the decomposition process. In general, the decomposition into a few layers can meet the requirement, and the decomposition operation can be terminated at any time.
(2) Signal processing
After the decomposition of the signal f is completed, the wavelet signal w is eliminated by threshold quantizationj-1The purpose of signal noise reduction can be achieved by using the partial frequency. Common threshold functions are hard threshold and soft threshold functions, etc. Due to the particularity of the ultrasonic defect echo signal, the place with small signal fluctuation is generally electronic white noise and belongs to a part needing to be filtered; the larger peaks in the signal are the potential defect echoes that should be preserved. So that the threshold value is usedWhen the method is used for processing wavelet coefficients, a more flexible strategy is needed. Aiming at the problems in the soft threshold function and the hard threshold function, a new threshold function is adopted:
Figure BDA0002568574100000081
wherein, λ is a threshold value,
Figure BDA0002568574100000082
for noise-reduced signals, wjFor signals before noise reduction processing, m and n are adjustable coefficients, and the threshold function can be approached to one direction between a soft threshold function and a hard threshold function by changing the assignment of m and n, which is different from the pure soft threshold function and the hard threshold function in the prior art. The threshold function has the advantage that the noise reduction capability of the threshold function can be optimized by adjusting parameters, so that the best noise reduction effect is achieved.
(3) Discrete wavelet transform reconstruction algorithm
Reconstruction is a process corresponding to decomposition, and is also divided into 3 steps: initialization, iteration and termination. After threshold processing is carried out on all decomposed wavelet parts, j processed wavelet signals w are obtained0，w1，……，wj-2，wj-1And an approximation signal f0. Unlike the down-sampling with step size 2 which is performed when the decomposed partial approximation signal passes through the low pass filter and the wavelet signal passes through the high pass filter, the up-sampling with step size 2 is performed in each process of reconstructing the signal. Thus, the signal f is obtained by a reconstruction function, and the detailed flow is shown in fig. 4.
S3: algorithm performance evaluation
In order to verify the advantages and feasibility of the method, an experimental sample piece containing prefabricated defects such as flat-bottom holes, air holes and cracks is prepared by adopting laser synchronous powder feeding additive manufacturing experimental equipment and used for acquiring an ultrasonic detection signal, and the specific conditions of the defects are shown in table 1.
TABLE 1 Preset Defect summary sheet
Defect number Details of the defect
1 Flat bottom hole of phi 0.5mm and depth of 10mm
2 Flat bottom hole of phi 0.5mm and depth 15mm
3 Flat bottom hole of phi 0.5 and depth 20mm
4 Pore type defect 1
5 Defect of pore type 2
6 Pore type defect 3
7 Crack(s)
And (3) uniformly setting the gain to be 40dB during ultrasonic detection, using the ultrasonic detection equipment shown in figure 1, using engine oil as a coupling agent, and detecting potential defects by using a direct contact method. And importing the detection data into Matlab software to generate an original waveform of the defect signal. When the amplitude of the defect echo wave is too small compared with that of the upper surface echo wave, the ultrasonic reflected wave on the lower surface of the test block can also interfere with the defect detection signal, and result comparison is not facilitated. The original signal and the non-defective signal are cancelled and the reflected wave signals of the upper and lower surfaces are removed to obtain an original processing signal, as shown in fig. 5, and then the processed signal is subjected to noise reduction processing.
The wavelet denoising algorithm herein was evaluated using Matlab software. Sym4 was chosen as the wavelet basis, the number of decomposition levels was 3, the tunable coefficient m was set to 1, and n was set to 2. And (3) performing noise reduction treatment on all 7 groups of defect data by adopting a minimax soft threshold method and a text method respectively. The signal-to-noise ratio (SNR) and the Root Mean Square Error (RMSE) are used as criteria for evaluating the noise reduction effect of the signal.
Figure BDA0002568574100000091
Figure BDA0002568574100000101
Wherein x (n) is the original processed signal,
Figure BDA0002568574100000102
is a noise-reduced signal.
The results of the noise reduction processing of the defect ultrasonic reflected wave signal are shown in fig. 6-12. Two noise reduction algorithms compare signal-to-noise ratio (SNR) to Root Mean Square Error (RMSE) after noise reduction to fig. 13, 14.
6-12, it can be seen that both wavelet noise reduction algorithms can filter the ultrasound reflected wave signal noise of typical defects to some extent, compared with the original processed signal waveform. The minimax soft threshold algorithm can effectively reduce noise of different types of defect noisy signals, but certain fluctuation exists in a defect-free part of signals. After the noise-containing signal is subjected to noise reduction by using the algorithm, the signal curve is smoother, the signal amplitude is not obviously weakened at the signal peak position representing the existence of the defect, and the defect detection rate can be effectively improved.
From fig. 13 and 14, it can be seen that the present algorithm is smaller than the minimax soft threshold method in terms of signal-to-noise ratio, and slightly larger in terms of root mean square error. The reason for this is that if the ultrasonic defect detection is regarded as a system feedback, the incident ultrasonic sound wave corresponds to the input signal of the system, the mathematical model of the internal tissue of the workpiece to be measured corresponds to the response function of the system, and the reflected ultrasonic wave received by the ultrasonic probe is the output signal of the system. Unlike the general case, the defective reflected wave useful in this signal is a noise portion thereof. In fig. 13, the signal-to-noise ratio of the signal under the algorithm is small, which proves that the defect reflected wave signal is strong; in fig. 14, the root mean square error of the signal is larger under the algorithm herein, demonstrating that the defect signal is more prominent than the signal noise floor. The two points illustrate that the noise reduction algorithm provided by the invention has better adaptability due to the adjusting coefficient, and is suitable for different types of defects (flat bottom holes, air holes and cracks).
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (8)

1. A noise reduction method of an ultrasonic detection signal based on an additive defect is characterized by comprising the following steps: offsetting the original signal with the defect and the signal without the defect and removing the reflected wave signals on the upper surface and the lower surface to obtain the ultrasonic reflected wave signals of the workpiece to be measured; and performing discrete wavelet denoising treatment on the ultrasonic reflected wave signal to improve the signal to noise ratio, so as to obtain an ultrasonic detection signal of the defect and identify the material increase defect.
2. The method for reducing the noise of the ultrasonic detection signal based on the additive defect according to claim 1, wherein the method is performed by using an ultrasonic detection system, the ultrasonic detection system comprises a phased array integrated instrument, an upper computer and an ultrasonic probe, the upper computer outputs a signal to enable the phased array integrated instrument to control the ultrasonic probe to emit and collect an original echo signal, and the upper computer processes the defect to identify the defect.
3. The method for denoising an ultrasonic detection signal based on an additive defect according to claim 1, wherein the discrete wavelet denoising process comprises:
discrete wavelet decomposition: finding the best approximation f of the ultrasonic reflected wave signal fjThen, down-sampling decomposition is carried out step by step until all wavelet components are obtained;
signal processing: eliminating partial frequency in wavelet component by threshold value quantization method to achieve the purpose of signal noise reduction;
and (3) discrete wavelet reconstruction: and performing up-sampling reconstruction on the wavelet component after signal processing to obtain a function to obtain an output signal of discrete wavelet de-noising processing, wherein the up-sampling reconstruction is an inverse process of discrete wavelet decomposition.
4. The method of claim 3, wherein the discrete wavelet decomposition comprises:
a. initialization: the purpose is to find the best approximation f of the ultrasonic reflected wave signal fjDue to fjFor signal f in approximate space VjOrthogonal projection P ofjf, when the scale function phi is tightly supported and the number of decomposition layers j is sufficiently large, there are:
Figure FDA0002568574090000011
Figure FDA0002568574090000021
where j is 1,2,3,4,5,6, m is an adjustment coefficient, m may be 1,2,3,4,5,6 … …, and the larger j is, the higher the approximation degree is, and f is, the larger f is, the higher f is, the lower the adjustment coefficient is, and the lower the adjustment coefficient is, the higher the adjustment coefficient is, and the lower the adjustment coefficient is, and the higher the adjustment coefficient isjThe accuracy of (f) depends on the smoothness of j and k, and f;
b. iteration: is to best approximate the signal fjCarrying out a downsampling decomposition process step by step;
for the best approximation signal fjDecomposed into the approximate part f of the next stage by a low-pass filterj-1The response of the low-pass filter is g [ n ]](ii) a Decomposition into wavelet portions w of the next stage by a high-pass filterj-1The response of the high-pass filter is h [ n ]]；
Figure FDA0002568574090000022
Figure FDA0002568574090000023
c. And (4) terminating: to fj-1The above decomposition process is repeated until all sample points are exhausted when j is 0.
5. The method of claim 3, wherein the signal processing comprises:
adopting the following threshold function to filter out electronic white noise and retain potential defect echo:
Figure FDA0002568574090000024
wherein, λ is a threshold value,
Figure FDA0002568574090000025
Figure FDA0002568574090000026
for noise-reduced signals, wjFor signals before noise reduction processing, m and n are adjustable coefficients, and the assignment of m and n is changed to enable the threshold function to approach to one direction between the soft threshold function and the hard threshold function.
6. The method for denoising ultrasonic detection signals based on additive defects according to claim 3, wherein the discrete wavelet signal reconstruction is: for signal processingAll j wavelet signals w obtained after the last step0，w1，……，wj-2，wj-1Using high-pass filtering, for an approximation signal f0The upsampling is performed using low pass filtering, resulting in the signal f.
7. The method for reducing noise based on the ultrasonic detection signal of the additive defect according to claim 1, wherein the method can be used for detecting laser additive products with flat-bottom holes, air holes and crack defects.
8. The method for reducing the noise of the ultrasonic detection signal based on the additive defect according to any one of claims 1 to 7, further comprising obtaining a waveform of the ultrasonic detection signal of the defect, and calculating a signal-to-noise ratio and a root-mean-square error index for comparison with a minimax soft threshold method to verify that the method is effective.
CN202010635245.7A 2020-07-03 2020-07-03 Noise reduction method of ultrasonic detection signal based on additive defect Active CN113884581B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010635245.7A CN113884581B (en) 2020-07-03 2020-07-03 Noise reduction method of ultrasonic detection signal based on additive defect

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010635245.7A CN113884581B (en) 2020-07-03 2020-07-03 Noise reduction method of ultrasonic detection signal based on additive defect

Publications (2)

Publication Number Publication Date
CN113884581A true CN113884581A (en) 2022-01-04
CN113884581B CN113884581B (en) 2022-08-30

Family

ID=79013239

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010635245.7A Active CN113884581B (en) 2020-07-03 2020-07-03 Noise reduction method of ultrasonic detection signal based on additive defect

Country Status (1)

Country Link
CN (1) CN113884581B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5841890A (en) * 1996-05-06 1998-11-24 Northrop Grumman Corporation Multi-dimensional wavelet tomography
US20040008904A1 (en) * 2003-07-10 2004-01-15 Samsung Electronics Co., Ltd. Method and apparatus for noise reduction using discrete wavelet transform
CN1808113A (en) * 2005-01-21 2006-07-26 宝山钢铁股份有限公司 Method for detecting inner defect of roller using ultrasonic wave
JP2011004345A (en) * 2009-06-22 2011-01-06 Nippon Hoso Kyokai <Nhk> Spatial downsampling processor, spatial upsampling processor, encoding apparatus, decoding apparatus, and program
CN105388212A (en) * 2015-10-23 2016-03-09 北京大学 Ultrasonic detecting method of defects in thick-wall composite tubular structure
CN105740762A (en) * 2016-01-13 2016-07-06 陈勇 Signal-noise separation optimization method
CN107576974A (en) * 2017-08-11 2018-01-12 东南大学 A kind of Multipath Errors extracting method based on adaptive semisoft shrinkage wavelet transformation

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5841890A (en) * 1996-05-06 1998-11-24 Northrop Grumman Corporation Multi-dimensional wavelet tomography
US20040008904A1 (en) * 2003-07-10 2004-01-15 Samsung Electronics Co., Ltd. Method and apparatus for noise reduction using discrete wavelet transform
CN1808113A (en) * 2005-01-21 2006-07-26 宝山钢铁股份有限公司 Method for detecting inner defect of roller using ultrasonic wave
JP2011004345A (en) * 2009-06-22 2011-01-06 Nippon Hoso Kyokai <Nhk> Spatial downsampling processor, spatial upsampling processor, encoding apparatus, decoding apparatus, and program
CN105388212A (en) * 2015-10-23 2016-03-09 北京大学 Ultrasonic detecting method of defects in thick-wall composite tubular structure
CN105740762A (en) * 2016-01-13 2016-07-06 陈勇 Signal-noise separation optimization method
CN107576974A (en) * 2017-08-11 2018-01-12 东南大学 A kind of Multipath Errors extracting method based on adaptive semisoft shrinkage wavelet transformation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
蒋鹏: "小波理论在信号去噪和数据压缩中的应用研究", 《中国博士学位论文全文数据库 (信息科技辑)》 *

Also Published As

Publication number Publication date
CN113884581B (en) 2022-08-30

Similar Documents

Publication Publication Date Title
CN109580787B (en) Ultrasonic echo denoising method for ultrasonic detection of transformer high-voltage bushing lead
CN103512960B (en) A kind of supersonic array formation method
CN112858183B (en) Additive manufacturing laser ultrasonic signal defect imaging method based on waveform separation
Hoseini et al. Denoising ultrasonic pulse-echo signal using two-dimensional analytic wavelet thresholding
CN111795931A (en) Reconstruction extraction method for laser ultrasonic defect detection diffraction echo signal
CN114264724A (en) Ultrasonic signal processing method based on CEEMDAN combined wavelet packet threshold
CN115840879A (en) Additive manufacturing component ultrasonic detection echo signal processing method and system
Giurgiutiu et al. Comparison of short-time fourier transform and wavelet transform of transient and tone burst wave propagation signals for structural health monitoring
CN113884581B (en) Noise reduction method of ultrasonic detection signal based on additive defect
CN111766306B (en) Full-focus imaging method and device
Mao et al. Noise suppression and flaw detection of ultrasonic signals via empirical mode decomposition
CN110780162B (en) Method for extracting partial discharge signal of primary and secondary fusion power distribution switch and detection device
CN112147226B (en) Optimal decomposition layer number determining method based on laser ultrasonic signal wavelet denoising
Guo et al. Reconstructing outside pass-band data to improve time resolution in ultrasonic detection
CN112183297B (en) Ultrasonic phased array signal sparse feature extraction method
CN110082436B (en) High-lift-off electromagnetic ultrasonic signal denoising method based on variational mode
Pardo et al. Shift invariant wavelet denoising of ultrasonic traces
Wang et al. An improved rail crack detection method based on the wavelet subband LMS adaptive filtering and wavelet analysis
CN116818903A (en) Defect detection method based on ultrasound
Yang et al. Matching-pursuit-based adaptive wavelet-packet atomic decomposition applied in ultrasonic inspection
Chen et al. Signal de-noising in ultrasonic testing based on stationary wavelet transform
Pipa et al. Adaptive time corrected gain for ultrasound through time-varying wiener deconvolution
Xu et al. A Threshold Denoising Method for Ultrasonic Signal Based on CEEMDAN and Spatial Correlation
CN112858182B (en) Laser ultrasonic signal defect echo signal extraction method based on waveform separation
Baradarani et al. Blind vision for real-time inspection of spot welds

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