CN110057918B - Method and system for quantitatively identifying damage of composite material under strong noise background - Google Patents
Method and system for quantitatively identifying damage of composite material under strong noise background Download PDFInfo
- Publication number
- CN110057918B CN110057918B CN201910458874.4A CN201910458874A CN110057918B CN 110057918 B CN110057918 B CN 110057918B CN 201910458874 A CN201910458874 A CN 201910458874A CN 110057918 B CN110057918 B CN 110057918B
- Authority
- CN
- China
- Prior art keywords
- damage
- strong noise
- composite material
- signals
- lamb 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.)
- Active
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
- G01N29/04—Analysing solids
- G01N29/12—Analysing solids by measuring frequency or resonance of acoustic waves
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a method and a system for quantitatively identifying damage of a composite material under a strong noise background, wherein the damage of different degrees is simulated on the composite material, different positions are changed, and Lamb wave response signals of different degrees and different positions are collected; adding a strong noise signal with a certain signal-to-noise ratio into the collected Lamb wave signals to simulate the Lamb wave signals collected under the background of strong noise; eliminating strong noise signals to obtain effective signals; dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction; and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information. The reliable positioning and accurate quantitative recognition of the composite material structure in a strong noise environment are realized.
Description
Technical Field
The disclosure belongs to the field of material damage information analysis, and relates to a method and a system for quantitatively identifying damage of a composite material under a strong noise background.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Carbon fiber Reinforced Composites (CFRP) play an important role in the aerospace industry due to their light weight, high strength, and high design ability. For example, CR929 aircraft have carbon fiber composite content as high as 50%. However, carbon fiber composite structures are susceptible to invisible damage caused by external impact and stress concentration during the manufacturing process or in-service application process, and even cause serious accidents. Therefore, to ensure the safety of the carbon fiber composite structure, a method for locating and quantifying the damage is needed.
Lamb waves are the focus of research on detection and evaluation of composite materials due to long propagation distance, low cost and good sensitivity to various defects. In order to determine the damage position of the composite material, researchers have studied methods such as a geometric positioning method and a time-of-flight method. The geometric positioning method and the time-of-flight method need to use the wave velocity to carry out position determination, however, due to the frequency dispersion characteristic of Lamb waves, the propagation speed is a function of frequency and material thickness, so that the wave velocity is not constant, and reliable positioning is difficult to realize. For quantitative damage identification, existing research extracts damage indicators from the amplitude, phase change and energy of Lamb wave signals to quantify the damage size. However, according to the knowledge of the inventor, because the propagation mechanism of Lamb waves in the composite material plate is not clear, the direct quantification of the damage degree by using the signal characteristic index is difficult. In addition, signals acquired by the sensor inevitably contain noise signals, and the wavelet transform is adopted by the increase of the roadside mobile units to eliminate the noise in the signals, so that conditions are provided for realizing damage positioning by adopting a time difference positioning method. Boudraa realizes signal denoising based on empirical mode decomposition, and relates to filtering or thresholding of intrinsic functions of each mode and reconstruction of an estimated signal by using the processed intrinsic functions of the modes. In addition, a fractal denoising method and a neural network denoising method are adopted, or a fractional differentiation method is adopted for signal denoising, so that the noise of the signal can be removed more effectively without signal priori knowledge, and the detail characteristics of the main signal are better reserved. However, these studies are mostly conducted in a laboratory environment, and the problem of strong noise interference such as noise generated by vibration of an aircraft wing structure and random noise (noise of a data acquisition system and interference of an external environment) in practical application is not considered.
In summary, when strong noise exists, weak differences between signals in a non-damage state are extremely easy to submerge, and damage assessment cannot be achieved by using the existing quantitative analysis method, namely the damage assessment of the existing composite material plate cannot be shifted to practical application from laboratory research, and the problem of damage feature extraction under the background of the strong noise must be solved.
Disclosure of Invention
The method and the system can quickly and accurately extract damage features under the environment with large data volume and strong noise, are used for positioning, detecting and quantitatively identifying the damage of the composite material plate, and overcome the defects that the traditional damage positioning method cannot reliably judge the damage position of the composite material based on Lamb wave velocity and cannot accurately and quantitatively identify the damage depending on signal features under the environment with strong noise.
According to some embodiments, the following technical scheme is adopted in the disclosure:
a method for quantitatively identifying damage of a composite material under a strong noise background comprises the following steps:
simulating damage of different degrees on the composite material, changing different positions, and collecting Lamb wave response signals of different degrees and different positions;
adding a strong noise signal with a certain signal-to-noise ratio into the collected Lamb wave signals to simulate the Lamb wave signals collected under the background of strong noise;
eliminating strong noise signals to obtain effective signals;
dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction;
and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information.
As a possible implementation mode, the structural strain field is changed by using the mass block to simulate real damage, damage of different degrees is simulated by changing different masses, and the position of the damage is changed by changing the arrangement position of the mass block.
As a possible implementation manner, the damage data acquisition system is used for acquiring damage data, and specifically comprises an arbitrary function generator, an amplifier, a plurality of piezoelectric sensors and an oscilloscope, wherein Lamb wave signals generated by the arbitrary function generator are amplified by the amplifier and loaded in at least one piezoelectric sensor, and the rest piezoelectric sensors acquire Lamb wave signals of mass blocks with different masses at different positions through the oscilloscope.
As a further limitation, multiple sets of data are collected, with each set of data collected multiple times.
As a possible implementation, the strong noise signal is rejected using a synchronous compression wavelet transform algorithm.
As a possible implementation mode, the effective lamb wave signals are subjected to Fourier transform, and the time domain signals are converted into frequency domain extraction features, so that the relationship between the change of frequency response and the degree and position of structural damage can be embodied.
As a further limitation, the self-encoder is trained layer by layer in a greedy learning manner and is formed by stacking the trained self-encoders.
By way of further limitation, the training process of the self-encoder includes two stages:
the first stage is as follows: inputting a sample into a first SAE network and fully training to obtain a parameter of a first layer, then taking the output of the first layer as the input of a next SAE, obtaining the parameter of the layer after the model is fully trained again, stacking the trained SAE models together, and repeating the steps until all SAEs are trained;
and a second stage: and adding a layer of neural network on the top layer, initializing the neural network by using the parameters learned in the first stage, and performing supervised fine tuning on each parameter obtained by training by using a back propagation algorithm.
A system for quantitatively identifying damage to a composite material on a strong noise background, comprising:
the acquisition system is configured to simulate different degrees of damage on the composite material, change different positions and acquire Lamb wave response signals of different degrees and different positions;
the signal processing system is configured to add a strong noise signal with a certain signal-to-noise ratio into the acquired Lamb wave signal to simulate the Lamb wave signal acquired under a strong noise background;
eliminating strong noise signals to obtain effective signals;
dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction;
and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information.
A computer readable storage medium, wherein a plurality of instructions are stored, said instructions are suitable for being loaded by a processor of a terminal device and executing said method for quantitative identification of composite material damage under a strong noise background.
A terminal device comprising a processor and a computer readable storage medium, the processor being configured to implement instructions; the computer readable storage medium is used for storing a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the composite material damage quantitative identification method under the strong noise background.
Compared with the prior art, the beneficial effect of this disclosure is:
the method effectively solves the problem that the damage of the composite material is difficult to judge and accurately and quantitatively identify the damage by using the characteristic index of the signal because the anisotropy of the composite material and the propagation of the wave in the composite material are not clear.
The method adopts the synchronous compression wavelet transform algorithm with strong noise removing capability to eliminate strong noise generated by the environment in the application environment and uses the SAE algorithm with strong nonlinear fitting capability to establish a damage identification model, thereby realizing reliable positioning and accurate quantitative identification of the composite material structure in the strong noise environment.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure and are not to limit the disclosure.
FIG. 1 is a schematic diagram of an actuator sensor arrangement;
FIG. 2 is a schematic flow chart of the method of the present embodiment;
FIG. 3 is a signal containing noise;
FIG. 4 is the signal after denoising;
FIG. 5 is a graph of the frequency spectrum of the response signals of three sensors with a 50g mass in position 1;
FIG. 6 is a graph of the frequency spectrum of the response signals of three sensors with a mass of 100g in position 1;
FIG. 7 is a graph of the frequency spectrum of the response signals of three sensors with a mass of 200g in position 1;
FIG. 8 is a graph of the frequency spectrum of the response signals of three sensors at position 25 for a 200g mass;
FIG. 9 is a graph of the frequency spectrum of the response signals of three sensors with a 200g mass at position 56;
FIG. 10 is a diagram of a self-encoder network architecture;
FIG. 11 is a diagram showing the results of lesion location determination and quantitative identification;
the specific implementation mode is as follows:
the present disclosure is further described with reference to the following drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present disclosure. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
A composite material damage positioning and quantitative identification method based on Lamb waves and an automatic encoder under a strong noise background comprises the following steps:
(1) a Lamb wave data acquisition system is built, and the Lamb wave data acquisition system comprises an arbitrary function generator, an amplifier, a piezoelectric sensor, an oscilloscope and a composite material plate, and is pasted with a square sensor array. One of the sensors sends lamb wave signals, and the other sensors acquire signals;
(2) the structural strain field is changed through the mass block to simulate real damage, different degrees of damage are simulated through changing different masses, and different positions are changed. The sensor collects Lamb wave response signals of different degrees and different positions;
(3) adding a strong noise signal with a signal-to-noise ratio of 3dB into the collected Lamb wave signal to simulate the collected Lamb wave signal under a strong noise background;
(4) the synchronous compression wavelet transformation algorithm realizes the noise elimination of the signals containing strong noise to obtain effective signals;
(5) the effective signal is divided into two parts, wherein one part is used as training data, the other part is used as test data, and Fourier transformation is carried out on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage, so that damage characteristic extraction is realized.
(6) And substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, and substituting the test data into the damage identification model for damage positioning and quantitative identification.
The signal-to-noise ratio calculation formula of the step (3) is as follows:
the synchronous compression wavelet transformation algorithm of the step (4) is as follows:
from the wavelet function, for function x (t) ε L2(R) continuous wavelet transform:
where ". x" is the conjugate of the function and W (a, b) is the wavelet transform coefficient.
Calculating instantaneous frequency omega by using obtained wavelet coefficientx(a, b) defined as:
this converts the time-scale plane (b, a) to a time-frequency plane (b, ω (a, b)), which can convert any frequency ω tolPeripheral zone
Is compressed to ωlIn this way, the value T (ω) of the synchronous compression transform can be obtainedlAnd b), thereby achieving the purpose of improving the time-frequency resolution. I.e. the synchronous compression transform can be expressed as:
wherein N represents
akAnd k is the discrete scale and the number of scales.
Lamb wave signal is discrete signal, and the above formula has medium scale coordinate delta akComprises the following steps: Δ ak＝ak-ak-1The frequency coordinate Δ ω is: Δ ω ═ ωl-ωl-1。
The inverse transform of the synchronous compression wavelet transform is:
in the formula (I), the compound is shown in the specification,
taking a finite value of psi*(xi) is the basic d xi wavelet function conjugate fourier transform.
The reconstructed signal can be obtained by the above equation.
And in SAE model training, damage positioning and quantitative identification, SAE training is carried out on the training data to obtain a composite plate damage position judgment and timing identification model, and the test data is substituted into the model to carry out result testing.
The self-encoder model in the step (6) is as follows:
the self-encoder trains the self-encoder layer by layer in a greedy learning mode and then is formed by stacking the trained self-encoders. The training process of the auto-encoder (SAE) is divided into two phases: unsupervised feature learning and supervised fine tuning.
The first stage is as follows: samples are input into the first SAE network and sufficiently trained to obtain a first layer parameter θ1Then, the output of the first layer is used as the input of the next SAE, and when the model is fully trained again, the parameter theta of the layer is obtained2. And stacking the trained SAE models together, and repeating the steps until all SAEs are trained.
And a second stage: and adding a layer of neural network on the top layer, initializing the neural network by using the parameters learned in the first stage, and performing supervised fine tuning on each parameter obtained by training by using a back propagation algorithm.
And the following product examples are provided:
a system for quantitatively identifying damage to a composite material on a strong noise background, comprising:
the acquisition system is configured to simulate different degrees of damage on the composite material, change different positions and acquire Lamb wave response signals of different degrees and different positions;
the signal processing system is configured to add a strong noise signal with a certain signal-to-noise ratio into the acquired Lamb wave signal to simulate the Lamb wave signal acquired under a strong noise background;
eliminating strong noise signals to obtain effective signals;
dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction;
and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information.
A computer readable storage medium, wherein a plurality of instructions are stored, said instructions are suitable for being loaded by a processor of a terminal device and executing said method for quantitative identification of composite material damage under a strong noise background.
A terminal device comprising a processor and a computer readable storage medium, the processor being configured to implement instructions; the computer readable storage medium is used for storing a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the composite material damage quantitative identification method under the strong noise background.
As a typical example, fig. 1 shows a carbon fiber composite material plate used in this example, which has a size of 60cm × 60cm, and 64 squares are uniformly drawn in the center of the plate, each square having a size of 3cm × 3 cm. In order to achieve the localization of the damage, 4 piezoelectric sensors are arranged in the plate, one of which transmits a Lamb wave signal at 50KHz as an exciter and the remaining 3 sensors receive response signals. The damage is realized by changing the strain field of the composite material structure by adopting 50g, 100g and 200g of mass blocks.
Fig. 2 is a schematic flow chart of the method of this embodiment, and a data acquisition system is first constructed, which includes an arbitrary function generator, an amplifier, a piezoelectric sensor, an oscilloscope, and a composite material plate, where a Lamb wave signal generated by the arbitrary function generator is amplified by the amplifier and loaded in the piezoelectric sensor, and the remaining sensors acquire Lamb wave signals at different positions of mass blocks of different masses through the oscilloscope, and each group acquires 150 times, where 64 × 3 is 192 groups in total. Adding a noise with a signal ratio of 3dB to these signals simulates a strong noise background, and fig. 3 is a signal containing strong noise. And eliminating strong noise by adopting synchronous compression wavelet transform, wherein a graph 4 shows that signals with strong noise are eliminated to obtain effective Lamb wave signals.
Because damage can cause the change of the dynamic characteristics of the structure, the lamb wave signals are subjected to Fourier transform, and the time domain signals are converted into frequency domain signals, so that the relationship between the change of frequency response and the degree and position of the structural damage can be embodied. Fig. 5 to 9 are frequency spectrum diagrams of the same mass at the same position and the same mass at different positions, which show that the damage degree is different and the damage position is different, so that a complex mapping relationship between the frequency spectrum and the damage can be established through an SAE algorithm.
In order to realize the training and testing of the model, 140 groups of random selection are selected from 150 groups of data to form training data for model training, and 1 group of random selection type testing data are selected from the remaining 10 groups for model testing. And putting the training data into an SAE algorithm for model training, wherein FIG. 10 is an SAE network structure diagram, and the network structure is set to [ 900100193 ] during training to obtain an SAE damage identification model. The test data is substituted into the model for testing, and the test result is shown in fig. 11, it can be seen that under 192 kinds of damage, only 1 kind of mass blocks are in error in position identification when the mass blocks are 200g, but the correct damage identification rate is 99.48% at the adjacent positions, and it can be seen that the method provided by the embodiment can well solve the problems of composite material structure damage position determination and damage degree quantitative identification under the strong noise background.
As will be appreciated by one skilled in the art, embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure, and various modifications and changes may be made to the present disclosure by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present disclosure should be included in the protection scope of the present disclosure.
Although the present disclosure has been described with reference to specific embodiments, it should be understood that the scope of the present disclosure is not limited thereto, and those skilled in the art will appreciate that various modifications and changes can be made without departing from the spirit and scope of the present disclosure.
Claims (9)
1. A method for quantitatively identifying damage of a composite material under a strong noise background is characterized by comprising the following steps: the method comprises the following steps:
simulating damage of different degrees on the composite material, changing different positions, utilizing the mass block to change a structural strain field to simulate real damage, simulating damage of different degrees by changing different masses, and changing the position of the damage by changing the arrangement position of the mass block;
collecting Lamb wave response signals of different degrees and different positions; adding a strong noise signal with a certain signal-to-noise ratio into the collected Lamb wave signals to simulate the Lamb wave signals collected under the background of strong noise;
eliminating noise signals containing strong noise signals to obtain effective signals;
dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction;
and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information.
2. The method for quantitatively identifying the damage of the composite material under the strong noise background as claimed in claim 1, wherein the method comprises the following steps: the method comprises the steps that damage data are collected by using a data collection system, and specifically comprises an arbitrary function generator, an amplifier, a plurality of piezoelectric sensors and an oscilloscope, wherein Lamb wave signals generated by the arbitrary function generator are amplified by the amplifier and loaded in at least one piezoelectric sensor, and the rest piezoelectric sensors collect Lamb wave signals of mass blocks with different masses at different positions through the oscilloscope.
3. The method for quantitatively identifying the damage of the composite material under the strong noise background as claimed in claim 1, wherein the method comprises the following steps: multiple sets of data are collected, and each set of data is collected for multiple times.
4. The method for quantitatively identifying the damage of the composite material under the strong noise background as claimed in claim 1, wherein the method comprises the following steps: and eliminating the noise signals containing strong noise signals by using a synchronous compression wavelet transform algorithm.
5. The method for quantitatively identifying the damage of the composite material under the strong noise background as claimed in claim 1, wherein the method comprises the following steps: the effective lamb wave signals are subjected to Fourier transform, and the time domain signals are converted into frequency domain signals, so that the relationship between the change of frequency response and the degree and position of structural damage can be embodied.
6. The method for quantitatively identifying the damage of the composite material under the strong noise background as claimed in claim 1, wherein the method comprises the following steps: the automatic encoder is trained layer by layer in a greedy learning mode and formed by stacking the trained automatic encoders;
or, the training process of the automatic encoder comprises two stages:
the first stage is as follows: inputting a sample into a first SAE network and fully training to obtain a parameter of a first layer, then taking the output of the first layer as the input of the next SAE, obtaining the parameter of the layer after the model is fully trained again, stacking the trained SAE models together, and repeating the steps until all SAEs are trained;
and a second stage: and adding a layer of neural network on the top layer, initializing the neural network by using the parameters learned in the first stage, and performing supervised fine tuning on each parameter obtained by training by using a back propagation algorithm.
7. A composite material damage quantitative identification system under a strong noise background is characterized in that: the method comprises the following steps:
the acquisition system is configured to simulate different degrees of damage on the composite material and change different positions, the structural strain field is changed by the mass block to simulate real damage, different degrees of damage are simulated by changing different masses, and the position of the damage is changed by changing the setting position of the mass block;
collecting Lamb wave response signals of different degrees and different positions;
the signal processing system is configured to add a strong noise signal with a certain signal-to-noise ratio into the acquired Lamb wave signal to simulate the Lamb wave signal acquired under a strong noise background;
eliminating strong noise signals to obtain effective signals;
dividing the effective signal into two parts, wherein one part is used as training data, the other part is used as test data, and performing Fourier transform on the training data and the test data to obtain frequency spectrum data corresponding to different degrees and different positions of damage so as to realize damage characteristic extraction;
and substituting the training data into the automatic encoder for training to obtain an automatic encoder damage identification model, substituting the test data into the trained damage identification model, and outputting according to the model to obtain damage positioning and quantitative identification information.
8. A computer-readable storage medium characterized by: stored with instructions adapted to be loaded by a processor of a terminal device and to perform a method for quantitative identification of damage to a composite material on a strong noise background according to any one of claims 1 to 6.
9. A terminal device is characterized in that: the system comprises a processor and a computer readable storage medium, wherein the processor is used for realizing instructions; the computer readable storage medium is used for storing a plurality of instructions, the instructions are suitable for being loaded by a processor and executing the composite material damage quantitative identification method under the strong noise background of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910458874.4A CN110057918B (en) | 2019-05-29 | 2019-05-29 | Method and system for quantitatively identifying damage of composite material under strong noise background |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910458874.4A CN110057918B (en) | 2019-05-29 | 2019-05-29 | Method and system for quantitatively identifying damage of composite material under strong noise background |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110057918A CN110057918A (en) | 2019-07-26 |
| CN110057918B true CN110057918B (en) | 2020-11-06 |
Family
ID=67325040
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910458874.4A Active CN110057918B (en) | 2019-05-29 | 2019-05-29 | Method and system for quantitatively identifying damage of composite material under strong noise background |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110057918B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110865123A (en) * | 2019-11-14 | 2020-03-06 | 天津大学 | Method for calculating laser surface acoustic wave frequency dispersion curve |
| CN112684012A (en) * | 2020-12-02 | 2021-04-20 | 青岛科技大学 | Equipment key force-bearing structural part fault diagnosis method based on multi-parameter information fusion |
| CN113740425B (en) * | 2021-07-12 | 2022-11-18 | 北京航空航天大学 | Composite material laminated structure anisotropic damage identification method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106198551A (en) * | 2016-08-01 | 2016-12-07 | 南方电网科学研究院有限责任公司 | The detection method of a kind of transmission line of electricity defect and device |
| CN107016241A (en) * | 2017-04-05 | 2017-08-04 | 重庆交通大学 | Based on the rotating machinery lifetime stage recognition methods for adding sample enhancing depth own coding learning network of making an uproar |
| CN107228942A (en) * | 2017-08-01 | 2017-10-03 | 福州大学 | Fluorescence immune chromatography detection method and device based on sparse own coding neutral net |
| CN108562709A (en) * | 2018-04-25 | 2018-09-21 | 重庆工商大学 | A kind of sewage disposal system water quality monitoring method for early warning based on convolution self-encoding encoder extreme learning machine |
| CN109490316A (en) * | 2018-11-30 | 2019-03-19 | 熵智科技(深圳)有限公司 | A kind of surface defects detection algorithm based on machine vision |
| CN109613178A (en) * | 2018-11-05 | 2019-04-12 | 广东奥博信息产业股份有限公司 | A kind of method and system based on recurrent neural networks prediction air pollution |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100580445C (en) * | 2007-10-12 | 2010-01-13 | 南京航空航天大学 | Engineering structure damage active monitoring Lamb wave time-reversal focusing method |
| CN101694484A (en) * | 2009-10-22 | 2010-04-14 | 中国第一重型机械股份公司 | Method for ultrasonic locating defect in austenitic stainless steel weld joint |
| CN203422355U (en) * | 2013-07-17 | 2014-02-05 | 浙江方圆金属材料检测有限公司 | Defect simulation and comparison test cube for ultrasonic detection |
| CN104297346A (en) * | 2014-09-11 | 2015-01-21 | 天津大学 | Nondestructive detection system of sheet metal by ultrasonic planar guided-wave and detection method thereof |
| CN106742057B (en) * | 2016-12-23 | 2019-05-07 | 湖南科技大学 | Aircraft skin damage monitoring device and method based on wireless piezoelectric sensing technology |
| CN109406629A (en) * | 2018-10-15 | 2019-03-01 | 成都飞机工业（集团）有限责任公司 | A kind of test block of the angle R and production method for composite structure ultrasound detection |
-
2019
- 2019-05-29 CN CN201910458874.4A patent/CN110057918B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106198551A (en) * | 2016-08-01 | 2016-12-07 | 南方电网科学研究院有限责任公司 | The detection method of a kind of transmission line of electricity defect and device |
| CN107016241A (en) * | 2017-04-05 | 2017-08-04 | 重庆交通大学 | Based on the rotating machinery lifetime stage recognition methods for adding sample enhancing depth own coding learning network of making an uproar |
| CN107228942A (en) * | 2017-08-01 | 2017-10-03 | 福州大学 | Fluorescence immune chromatography detection method and device based on sparse own coding neutral net |
| CN108562709A (en) * | 2018-04-25 | 2018-09-21 | 重庆工商大学 | A kind of sewage disposal system water quality monitoring method for early warning based on convolution self-encoding encoder extreme learning machine |
| CN109613178A (en) * | 2018-11-05 | 2019-04-12 | 广东奥博信息产业股份有限公司 | A kind of method and system based on recurrent neural networks prediction air pollution |
| CN109490316A (en) * | 2018-11-30 | 2019-03-19 | 熵智科技(深圳)有限公司 | A kind of surface defects detection algorithm based on machine vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110057918A (en) | 2019-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109682561B (en) | Method for automatically detecting free vibration response of high-speed railway bridge to identify mode | |
| CN110057918B (en) | Method and system for quantitatively identifying damage of composite material under strong noise background | |
| CN101836100B (en) | A method for detection and automatic identification of damage to rolling bearings | |
| CN109084186B (en) | Pipeline leakage signal identification method based on improved ELMD (ensemble empirical mode decomposition) multi-scale entropy | |
| CN107066736B (en) | Modal analysis and structural impact monitoring method based on compressive sampling | |
| CN105929024A (en) | Intelligent detection and quantitative recognition method for defect of concrete | |
| CN101865789A (en) | Fault detecting device of near field acoustic holography sound image mode identification and detecting method thereof | |
| Sen et al. | Sparsity-based approaches for damage detection in plates | |
| CN106289947B (en) | Method for identifying structural damage of light high-strength beam | |
| CN105488520A (en) | Multi-resolution singular-spectrum entropy and SVM based leakage acoustic emission signal identification method | |
| Li et al. | Acoustic emission signal source localization on plywood surface with cross-correlation method | |
| Feng et al. | Localization of low-velocity impact in CFRP plate using time–frequency features of guided wave and convolutional neural network | |
| Ma et al. | Two-stage damage identification based on modal strain energy and revised particle swarm optimization | |
| CN112285651A (en) | Method and system for identifying position of composite impact source based on feature extraction matrix similarity | |
| CN205749240U (en) | A kind of macromolecular material internal flaw the cannot-harm-detection device | |
| Jiang et al. | A hybrid multiple damages detection method for plate structures | |
| CN113627048B (en) | Method for quickly identifying structural damage based on local transfer rate function and pattern matching | |
| Saechai et al. | Test system for defect detection in construction materials with ultrasonic waves by support vector machine and neural network | |
| CN109801401A (en) | Fault detection method, computer-readable medium, fault detection means and ship | |
| Xu et al. | Determination of impact events on a plate-like composite structure | |
| Zhao et al. | Damage location detection of the CFRP composite plate based on neural network regression | |
| Al-hababi et al. | Time-frequency domain methods for the identification of breathing cracks in beam-like structures | |
| Wang et al. | Comparison of sparse representation and fourier discriminant methods: damage location classification in indirect lab-scale bridge structural health monitoring | |
| CN113640394A (en) | Fatigue crack identification method and system | |
| CN112131781A (en) | Steel structure damage detection method based on full-connection neural network and transfer rate function |
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 |