CN117191956A - Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus - Google Patents
Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus Download PDFInfo
- Publication number
- CN117191956A CN117191956A CN202311035526.9A CN202311035526A CN117191956A CN 117191956 A CN117191956 A CN 117191956A CN 202311035526 A CN202311035526 A CN 202311035526A CN 117191956 A CN117191956 A CN 117191956A
- Authority
- CN
- China
- Prior art keywords
- stress corrosion
- signal data
- waveform signal
- acoustic emission
- titanium alloy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000007797 corrosion Effects 0.000 title claims abstract description 162
- 238000005260 corrosion Methods 0.000 title claims abstract description 162
- 229910001069 Ti alloy Inorganic materials 0.000 title claims abstract description 75
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000003066 decision tree Methods 0.000 claims abstract description 41
- 238000013145 classification model Methods 0.000 claims abstract description 25
- 230000008569 process Effects 0.000 claims abstract description 22
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 239000000523 sample Substances 0.000 claims description 39
- 238000002474 experimental method Methods 0.000 claims description 34
- 238000012360 testing method Methods 0.000 claims description 27
- 239000000463 material Substances 0.000 claims description 25
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 7
- 238000011065 in-situ storage Methods 0.000 claims description 7
- 238000012512 characterization method Methods 0.000 claims description 6
- 239000007822 coupling agent Substances 0.000 claims description 6
- 229940099259 vaseline Drugs 0.000 claims description 6
- 239000011248 coating agent Substances 0.000 claims description 5
- 238000000576 coating method Methods 0.000 claims description 5
- 238000002372 labelling Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 abstract description 14
- 238000005516 engineering process Methods 0.000 description 13
- 238000012549 training Methods 0.000 description 11
- 230000007547 defect Effects 0.000 description 9
- 230000015654 memory Effects 0.000 description 6
- 239000000956 alloy Substances 0.000 description 5
- 238000005336 cracking Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 229910052751 metal Inorganic materials 0.000 description 5
- 239000002184 metal Substances 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 238000009659 non-destructive testing Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000013078 crystal Substances 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000000977 initiatory effect Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 229910045601 alloy Inorganic materials 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000003776 cleavage reaction Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 229910052739 hydrogen Inorganic materials 0.000 description 1
- 239000001257 hydrogen Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011081 inoculation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000007017 scission Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000010935 stainless steel Substances 0.000 description 1
- 229910001220 stainless steel Inorganic materials 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Abstract
The invention discloses a titanium alloy stress corrosion damage classification method and device based on acoustic emission, and relates to the technical field of titanium alloy corrosion detection. Comprising the following steps: acquiring stress corrosion damage signals to be classified by using an acoustic emission sensor; inputting stress corrosion damage signals into the constructed classification model; and obtaining a classification result of the stress corrosion damage signal according to the stress corrosion damage signal and the classification model, and obtaining the cause of the stress corrosion damage of the titanium alloy according to the classification result. The method can utilize the stress corrosion damage signal monitored by the acoustic emission sensor to carry out waveform classification through the decision tree algorithm so as to judge the cause of corrosion damage in the titanium alloy stress corrosion process.
Description
Technical Field
The invention relates to the technical field of titanium alloy corrosion detection, in particular to a titanium alloy stress corrosion damage classification method and device based on acoustic emission.
Background
The corrosion evaluation of titanium alloy in the deep open sea environment is a systematic engineering, and relates to complex scientific and technical problems such as evolution behavior of materials in a multi-factor coupling environment under service conditions, scale correlation from microcosmic to real structure and the like. The material corrosion data is the basis for researching the material corrosion failure rule and the material performance evolution. In recent years, with rapid development of data-driven modeling ideas, development of new generation artificial intelligence technology represented by machine learning provides a new idea for excavating material failure rules, predicting material corrosion rates and early warning failure boundaries in deep open sea environments. Because of numerous factors influencing corrosion of deep-open sea titanium alloy, the action rule is relatively complex, high nonlinearity exists, and the black box property of the deep learning model makes the effect of the deep learning model in research of corrosion rule of deep-open sea materials very limited.
Stress corrosion refers to the process of material failure of a metal by the combined action of strain and corrosion caused by residual or applied stress in the corrosive medium. Such corrosion generally penetrates the grains, so-called transcrystalline corrosion, while there are already conditions of intergranular corrosion. When the metal is subjected to stress corrosion, cracks from the outside to the inside only occur in local areas. The crack propagates while the trunk crack propagates, and several branches develop simultaneously. The trend of the crack is macroscopically perpendicular to the direction of the tensile stress, the microscopic fracture mechanism is generally along-crystal fracture, and also can be through-crystal cleavage fracture or a mixture of the two, and corrosion products and corrosion pits of mud patterns can be seen on the fracture surface. Stress corrosion in general is allBrittle fracture comprises an inoculation area, an expansion area and a transient fracture area. The crack growth rate of stress corrosion is generally 10 -6 ～10 -3 mm/min。
Acoustic emission technology is an emerging non-destructive testing technology. The acoustic emission phenomenon can be generated when the material is damaged such as pitting corrosion, bubbles, crack growth and the like in the corrosion process. Therefore, the technology can judge the corrosion damage and crack growth of the material by analyzing the collected acoustic emission signals. The acoustic emission technology has the greatest advantages that long-time real-time on-line detection and monitoring alarm can be realized, and no requirement is imposed on the shape and size of the detected object. In addition, the acoustic emission technology can detect the stress corrosion cracking process of stainless steel, the acoustic emission probe has high sensitivity and high response speed, and acoustic emission signals generated by crack initiation in stress corrosion can be collected. Because the acoustic emission phenomenon is ubiquitous, the detection sensitivity is high, and the detection process is easily interfered by external environment noise, so that the analysis and judgment of the detection result are affected. Therefore, noise factors should be eliminated as much as possible in acoustic emission detection. Acoustic emission is a dynamic process, so acoustic emission detection can only detect dynamic defects, which is a passive non-destructive detection technique. Analysis and evaluation of test results sometimes requires verification by other test methods. In the aspect of corrosion detection, the acoustic emission technology is mainly used for detecting stress corrosion cracking of materials, and the acoustic emission sensor has very high sensitivity, so that the technology can effectively detect crack initiation states in the stress corrosion process of the materials.
From the acoustic emission signals of microscopic deformations and crack release of the material, dynamic information of defects or cracks can be deduced. AE (Acoustic Emission ) is the point at which strain energy or elastic waves are released when the material is damaged. Acoustic emission phenomena can occur during the initial stages of material damage. The existence, the position, the expansion trend and the severity of the damage or the defect can be effectively discovered by timely capturing and collecting the acoustic emission signals.
The acoustic emission detection can realize qualitative, quantitative and positioning research of structural damage or defect, namely: judging the existence of defects and the types of defect damage, evaluating the severity of the defect damage, and determining the specific position of an acoustic emission source. When the sound source signal cannot be judged and disputed, other methods should be adopted for rechecking and verification. Therefore, acoustic emission technology is not an alternative to the original traditional nondestructive testing technology, but is a complement and improvement to the nondestructive testing technology.
Because the deep open sea environment titanium alloy corrosion basic data is less in accumulation but more in data sources, and strong heterogeneity exists among multisource experimental data, the data can be classified into sparse multisource heterogeneous characteristics, so that on one hand, the missing value in filling experimental data needs to be estimated to realize data enhancement, and on the other hand, the multisource heterogeneous data needs to be effectively integrated to realize complementation and correction of the data. And registering and correlating various monitoring data of the titanium alloy material under single-factor and multi-factor change conditions by utilizing basic data, physical test data, real sea test data, simulation analysis data and the like of the titanium alloy material in a deep open sea environment, so as to provide a foundation for subsequent data fusion and data modeling.
However, in the prior art, the types of stress corrosion damage of the titanium alloy cannot be accurately distinguished according to the existing corrosion data under the condition of no damage. By utilizing the technology to monitor the equipment or the component, the material defect type can be effectively distinguished, the corrosion stage of the material defect type can be determined, and technical support is provided for stress corrosion cracking detection of the metal component.
Disclosure of Invention
The invention provides the method for accurately distinguishing the types of the stress corrosion damage of the titanium alloy according to the existing corrosion data under the condition of no damage.
In order to solve the technical problems, the invention provides the following technical scheme:
in one aspect, the invention provides a titanium alloy stress corrosion damage classification method based on acoustic emission, which is realized by electronic equipment, and comprises the following steps:
s1, acquiring stress corrosion damage signals to be classified by utilizing an acoustic emission sensor.
S2, inputting stress corrosion damage signals into the constructed classification model.
S3, according to the stress corrosion damage signals and the classification model, obtaining classification results of the stress corrosion damage signals, and according to the classification results, obtaining the reasons of the stress corrosion damage of the titanium alloy.
Optionally, the process of constructing the classification model in S2 includes:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set.
S22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data.
S23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and the decision tree algorithm.
Optionally, the titanium alloy stress corrosion experiment with acoustic emission in S21 includes:
s211, fixing the stress corrosion test sample in a clamping block on a base of the material deformation in-situ characterization system.
S212, fixing the acoustic emission sensor serving as an experimental probe on the surface of the stress corrosion sample.
And S213, uniformly coating Vaseline coupling agent on the contact part of the experimental probe and the stress corrosion sample, and carrying out a titanium alloy stress corrosion experiment carrying acoustic emission.
Optionally, the acoustic emission sensor comprises: broadband sensor, preamplifier and acquisition device PCI-2.
Optionally, the extracting in S22 the waveform signal data set according to the experimental data set includes:
s221, extracting waveform signal data of voltage versus time at each impact from the experimental data set.
S222, preprocessing waveform signal data.
S223, extracting the preset number of waveform signal data from the preprocessed waveform signal data, performing waveform drawing on the extracted waveform signal data to obtain a drawing result, and labeling the waveform signal data according to the drawing result.
Optionally, preprocessing waveform signal data in S222 includes:
and acquiring test time in experimental information in the titanium alloy stress corrosion experimental process, and denoising and filtering waveform signal data.
And sorting the processed waveform signal data by using codes, naming the file name of the sorted waveform signal data as test time, and obtaining the file content of the waveform signal data from which the experimental information is removed.
Optionally, the tag in S223 includes: burst signals, continuous signals, and noise.
Optionally, the parameter setting of the decision tree algorithm in S23 includes:
the division Criterion of the decision tree is set to the base coefficient gini.
The maximum depth of the decision tree is set to None.
The minimum number of samples on the leaf node of the decision tree is set to 1.
In another aspect, the present invention provides an acoustic emission-based titanium alloy stress corrosion damage classification device, which is applied to implement an acoustic emission-based titanium alloy stress corrosion damage classification method, and the device includes:
the acquisition module is used for acquiring stress corrosion damage signals to be classified by utilizing the acoustic emission sensor.
And the input module is used for inputting the stress corrosion damage signals into the constructed classification model.
And the output module is used for obtaining a classification result of the stress corrosion damage signal according to the stress corrosion damage signal and the classification model and obtaining the cause of the stress corrosion damage of the titanium alloy according to the classification result.
Optionally, the input module is further configured to:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set.
S22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data.
S23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and the decision tree algorithm.
Optionally, the input module is further configured to:
s211, fixing the stress corrosion test sample in a clamping block on a base of the material deformation in-situ characterization system.
S212, fixing the acoustic emission sensor serving as an experimental probe on the surface of the stress corrosion sample.
And S213, uniformly coating Vaseline coupling agent on the contact part of the experimental probe and the stress corrosion sample, and carrying out a titanium alloy stress corrosion experiment carrying acoustic emission.
Optionally, the acoustic emission sensor comprises: broadband sensor, preamplifier and acquisition device PCI-2.
Optionally, the input module is further configured to:
s221, extracting waveform signal data of voltage versus time at each impact from the experimental data set.
S222, preprocessing waveform signal data.
S223, extracting the preset number of waveform signal data from the preprocessed waveform signal data, performing waveform drawing on the extracted waveform signal data to obtain a drawing result, and labeling the waveform signal data according to the drawing result.
Optionally, the input module is further configured to:
and acquiring test time in experimental information in the titanium alloy stress corrosion experimental process, and denoising and filtering waveform signal data.
And sorting the processed waveform signal data by using codes, naming the file name of the sorted waveform signal data as test time, and obtaining the file content of the waveform signal data from which the experimental information is removed.
Optionally, the tag comprises: burst signals, continuous signals, and noise.
Optionally, the input module is further configured to:
the division Criterion of the decision tree is set to the base coefficient gini.
The maximum depth of the decision tree is set to None.
The minimum number of samples on the leaf node of the decision tree is set to 1.
In one aspect, an electronic device is provided, the electronic device comprising a processor and a memory, the memory storing at least one instruction, the at least one instruction being loaded and executed by the processor to implement the acoustic emission based titanium alloy stress corrosion damage classification method described above.
In one aspect, a computer readable storage medium having stored therein at least one instruction loaded and executed by a processor to implement the above-described acoustic emission based titanium alloy stress corrosion damage classification method is provided.
Compared with the prior art, the technical scheme has at least the following beneficial effects:
according to the scheme, a titanium alloy stress corrosion experiment carrying acoustic emission is carried out, and an experiment data set is obtained; extracting a plurality of waveform signal data sets of voltage (millivolts) versus time (seconds) for each impact from the experimental data; waveform drawing is carried out on a part of signal data files from a large amount of signal data according to experimental conditions, and the signal data are labeled according to waveform patterns; taking the marked data as a training set, and classifying the unmarked signal data by adopting a decision tree algorithm to obtain waveform classification; and comparing the classified waveforms with an actual experiment to obtain the final cause of the stress corrosion damage of the titanium alloy. By adopting the method, the stress corrosion damage signal monitored by the acoustic emission sensor can be utilized, and the waveform classification is carried out through the decision tree algorithm, so that the reason of corrosion damage in the titanium alloy stress corrosion process is judged.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a titanium alloy stress corrosion damage classification method based on acoustic emission provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of a design of a titanium alloy stress corrosion experiment with acoustic emission provided in an embodiment of the present invention;
FIG. 3 is a second schematic diagram of a design of a titanium alloy stress corrosion experiment with acoustic emission provided in an embodiment of the present invention;
FIG. 4 is a block diagram of a titanium alloy stress corrosion damage classification device based on acoustic emission provided by an embodiment of the invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without creative efforts, based on the described embodiments of the present invention fall within the protection scope of the present invention.
As shown in fig. 1, the embodiment of the invention provides a titanium alloy stress corrosion damage classification method based on acoustic emission, which can be realized by electronic equipment. The flow chart of the titanium alloy stress corrosion damage classification method based on acoustic emission as shown in fig. 1, the processing flow of the method can comprise the following steps:
s1, acquiring stress corrosion damage signals to be classified by utilizing an acoustic emission sensor.
S2, inputting stress corrosion damage signals into the constructed classification model.
Optionally, the process of constructing the classification model in S2 includes S21-S23:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set. Optionally, the acoustic emission-carrying titanium alloy stress corrosion test in S21 includes S211-S213:
s211, fixing the stress corrosion test sample in a clamping block on a base of the material deformation in-situ characterization system.
S212, fixing the acoustic emission sensor serving as an experimental probe on the surface of the stress corrosion sample.
And S213, uniformly coating Vaseline coupling agent on the contact part of the experimental probe and the stress corrosion sample, and carrying out a titanium alloy stress corrosion experiment carrying acoustic emission.
In one possible embodiment, the stress corrosion coupon is fixed in the clamp block on the base of the material deformation in situ characterization system prior to the experiment in the manner shown in fig. 2 and 3. Then a piezoelectric narrowband resonant acoustic emission sensor with the model of R15a is used as an experimental probe and is fixed on the surface of a sample through a fixer so as to receive AE waves generated by stress corrosion of the sample, and when the probe is fixed, vaseline coupling agent is uniformly smeared on the contact part of the probe and the sample, so that acoustic emission signals generated by the sample in the stress corrosion test process can be well received by an amplifier.
The acoustic emission instrument consists of a broadband sensor of the American Physical Acoustic Co (PAC), a preamplifier and an acquisition device (PCI-2).
Further, AE signals are transmitted to the acquisition card through a 2/4/6 type preamplifier, the 2/4/6 type preamplifier is arranged at 40dB, and the attaching position of the acoustic emission probe and the fixed sample are shown in FIG. 2. In addition, the sample fixing base is connected with the back shell of the acoustic emission host through the copper wire so as to eliminate environmental noise. Then through a lead breaking test, namely an HB pencil with the diameter of 05mm is used, the pencil core stretches out by about 2.5mm, the included angle between the pencil core and the surface of a test piece is 30 degrees, the pencil core is slowly pressed to break, the pencil core is required to be broken for 3-4 times, and the emitted acoustic emission signals are collected. The fracture of the pencil lead is used as a simulated pulse sound source to simulate signals generated by structural or material deformation and fracture, the signals are used for determining threshold values and other parameters during detection, and finally 40dB threshold values and band-pass filtering between 20kHZ and 1000kHz are set.
S22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data.
Optionally, the extracting in S22 the waveform signal data set according to the experimental data set includes S221-S223:
s221, extracting waveform signal data of voltage versus time at each impact from the experimental data set.
In a possible implementation mode, the acoustic emission instrument is loaded while the titanium alloy material is subjected to stress corrosion, acoustic emission signals generated by the titanium alloy material are monitored simultaneously in the corrosion experiment process, and signal data display of 'impact' parameters are selected from acoustic emission parameters.
The signal data at each impact occurrence shown in the acoustic emission instrument was derived throughout the experiment, with the waveform data being voltage (millivolts) versus time (seconds) for each impact.
S222, preprocessing waveform signal data.
Optionally, preprocessing the waveform signal data in S222 includes S2221-S2222:
s2221, obtaining test time in experimental information in the titanium alloy stress corrosion experimental process, and carrying out denoising and filtering treatment on waveform signal data.
In a possible embodiment, the effective information in each txt signal file is extracted, i.e. the TEST TIME (TIME OF TEST) in the experimental information, i.e. the TIME point OF the change OF the metal crack during the stress corrosion is represented.
S2222, sorting the processed waveform signal data by using codes, naming the file name of the sorted waveform signal data as test time, and the file content is the waveform signal data from which the experimental information is removed.
In one possible implementation, the scrambled data is sorted into a form that can be used for analysis using code.
Specifically, it is necessary to extract the data in each txt file and perform denoising and filtering processes to remove the interference signal and noise. Then, waveform data of voltage (millivolts) versus time (seconds) of each impact can be organized into a file, the file name is test time, and the file content is pure acoustic emission signal data after experimental information is removed.
The specific code for processing the data is as follows:
s223, extracting the preset number of waveform signal data from the preprocessed waveform signal data, performing waveform drawing on the extracted waveform signal data to obtain a drawing result, and labeling the waveform signal data according to the drawing result.
Wherein, the label includes: burst signals, continuous signals, and noise.
In a possible implementation mode, part of the signal data files are taken from a large amount of signal data according to experimental conditions to carry out waveform drawing, and the signal data are labeled according to waveform patterns.
S23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and the decision tree algorithm.
Optionally, the parameter setting of the decision tree algorithm in S23 includes S231-S233:
s231, setting a division standard Criterion of the decision tree as a base coefficient gini.
S232, setting the maximum depth max_depth of the decision tree as None.
S233, setting the minimum sample number min_samples_leaf on the leaf node of the decision tree to be 1.
In a feasible implementation mode, marked data is used as a training set, and a decision tree algorithm is adopted to classify unmarked signal data to obtain waveform classification; and comparing the classified waveforms with an actual experiment to obtain the final cause of the stress corrosion damage of the titanium alloy.
Further, the basic idea of decision trees is to construct a tree with training data, each node representing a feature, each branch representing a value of the feature, and each leaf node representing a classification or regression result. By selecting and splitting features, the decision tree can progressively divide the dataset into smaller subsets until the leaf nodes are finally obtained, thereby completing the classification task.
B1, firstly, the data set is required to be divided into a training set and a testing set, wherein the training set is used for constructing a decision tree model, and the testing set is used for evaluating the performance of the model. On the training set, a suitable feature needs to be selected as the root node, and the data set is divided into a plurality of subsets, and each subset corresponds to one branch. The method of selecting the features may use information entropy, a keni index, or the like, in order to maximize the purity of each node. Specifically, the calculation formula of the information entropy is:
wherein X represents a random variable, n represents the number of X values, p i The probability that X takes a value i is indicated. The smaller the entropy of the information, the higher the purity of the data set.
The formula for calculating the base index is:
wherein X also represents a random variable, n represents the number of values of X, p i The probability that X takes a value i is indicated. The smaller the base index, the higher the purity of the data set.
B2, recursively repeating the step B1 for each subset until the data in all subsets belong to the same category or a certain stop condition is met. The constructed decision tree can be used to classify new unknown data.
Specifically, the classification process of the decision tree can be expressed by the following formula:
wherein y represents the classification result, x represents the input feature vector, w i Representing the weight of the ith node, g i (x) The determination condition of the i-th node is indicated. In the decision tree, each node corresponds to a feature, each branch corresponds to a value, and each leaf node corresponds to a category.
When training the decision tree model, proper features and decision conditions need to be selected to maximize the accuracy of classification. Common feature selection methods include information entropy and a base index, both of which aim to maximize the purity of each node. In classification, the input feature vectors are judged layer by layer from the root node until the leaf nodes are reached and the classification result is returned.
Training results show that the decision tree algorithm can effectively divide out-of-order and chapter-free acoustic emission signals into 3 different categories, and match and correspond the signal categories with actual damaged categories of the titanium alloy, wherein: the burst waveform is a continuous signal due to crack propagation of the sample, and is a noise-free disturbance signal due to plastic deformation of the crack tip of the sample.
In the embodiment, for a class of service materials of titanium alloy, performing a titanium alloy stress corrosion experiment carrying acoustic emission to obtain an experiment data set; extracting a plurality of waveform signal data sets of voltage (millivolts) versus time (seconds) for each impact from the experimental data; waveform drawing is carried out on a part of signal data files from a large amount of signal data according to experimental conditions, and the signal data are labeled according to waveform patterns; taking the marked data as a training set, and classifying the unmarked signal data by adopting a decision tree algorithm to obtain waveform classification; and comparing the classified waveforms with an actual experiment to obtain the final cause of the stress corrosion damage of the titanium alloy.
According to the titanium alloy stress corrosion damage classification method based on acoustic emission, an experimental system for monitoring stress corrosion based on acoustic emission detection technology is built, and stress corrosion cracking damage of the hydrogen-filled TC4 alloy is monitored. Acoustic emission signals are generated at each stage of stress corrosion cracking, and the damage condition of the component is analyzed by utilizing the impact signal waveform of each moment of the acoustic emission signals, so that a basis is provided for corrosion failure monitoring of the metal component. In addition, in-situ information of corrosion failure can be obtained while monitoring the component, and important reference and theoretical basis are provided for explaining stress corrosion mechanism.
S3, according to the stress corrosion damage signals and the classification model, obtaining classification results of the stress corrosion damage signals, and according to the classification results, obtaining the reasons of the stress corrosion damage of the titanium alloy.
In the embodiment of the invention, a titanium alloy stress corrosion experiment carrying acoustic emission is carried out, and an experimental data set is obtained; extracting a plurality of waveform signal data sets of voltage (millivolts) versus time (seconds) for each impact from the experimental data; waveform drawing is carried out on a part of signal data files from a large amount of signal data according to experimental conditions, and the signal data are labeled according to waveform patterns; taking the marked data as a training set, and classifying the unmarked signal data by adopting a decision tree algorithm to obtain waveform classification; and comparing the classified waveforms with an actual experiment to obtain the final cause of the stress corrosion damage of the titanium alloy. By adopting the method, the stress corrosion damage signal monitored by the acoustic emission sensor can be utilized, and the waveform classification is carried out through the decision tree algorithm, so that the reason of corrosion damage in the titanium alloy stress corrosion process is judged.
As shown in fig. 4, an embodiment of the present invention provides a titanium alloy stress corrosion damage classification device 400 based on acoustic emission, where the device 400 is applied to implement a titanium alloy stress corrosion damage classification method based on acoustic emission, and the device 400 includes:
the acquisition module 410 is configured to acquire a stress corrosion damage signal to be classified by using the acoustic emission sensor.
The input module 420 is configured to input the stress corrosion damage signal into the constructed classification model.
And the output module 430 is configured to obtain a classification result of the stress corrosion damage signal according to the stress corrosion damage signal and the classification model, and obtain a cause of the stress corrosion damage of the titanium alloy according to the classification result.
Optionally, the input module 420 is further configured to:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set.
S22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data.
S23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and the decision tree algorithm.
Optionally, the input module 420 is further configured to:
s211, fixing the stress corrosion test sample in a clamping block on a base of the material deformation in-situ characterization system.
S212, fixing the acoustic emission sensor serving as an experimental probe on the surface of the stress corrosion sample.
And S213, uniformly coating Vaseline coupling agent on the contact part of the experimental probe and the stress corrosion sample, and carrying out a titanium alloy stress corrosion experiment carrying acoustic emission.
Optionally, the acoustic emission sensor comprises: broadband sensor, preamplifier and acquisition device PCI-2.
Optionally, the input module 420 is further configured to:
s221, extracting waveform signal data of voltage versus time at each impact from the experimental data set.
S222, preprocessing waveform signal data.
S223, extracting the preset number of waveform signal data from the preprocessed waveform signal data, performing waveform drawing on the extracted waveform signal data to obtain a drawing result, and labeling the waveform signal data according to the drawing result.
Optionally, the input module 420 is further configured to:
and acquiring test time in experimental information in the titanium alloy stress corrosion experimental process, and denoising and filtering waveform signal data.
And sorting the processed waveform signal data by using codes, naming the file name of the sorted waveform signal data as test time, and obtaining the file content of the waveform signal data from which the experimental information is removed.
Optionally, the tag comprises: burst signals, continuous signals, and noise.
Optionally, the input module 420 is further configured to:
the division Criterion of the decision tree is set to the base coefficient gini.
The maximum depth of the decision tree is set to None.
The minimum number of samples on the leaf node of the decision tree is set to 1.
In the embodiment of the invention, a titanium alloy stress corrosion experiment carrying acoustic emission is carried out, and an experimental data set is obtained; extracting a plurality of waveform signal data sets of voltage (millivolts) versus time (seconds) for each impact from the experimental data; waveform drawing is carried out on a part of signal data files from a large amount of signal data according to experimental conditions, and the signal data are labeled according to waveform patterns; taking the marked data as a training set, and classifying the unmarked signal data by adopting a decision tree algorithm to obtain waveform classification; and comparing the classified waveforms with an actual experiment to obtain the final cause of the stress corrosion damage of the titanium alloy. By adopting the method, the stress corrosion damage signal monitored by the acoustic emission sensor can be utilized, and the waveform classification is carried out through the decision tree algorithm, so that the reason of corrosion damage in the titanium alloy stress corrosion process is judged.
Fig. 5 is a schematic structural diagram of an electronic device 500 according to an embodiment of the present invention, where the electronic device 500 may have a relatively large difference due to different configurations or performances, and may include one or more processors (central processing units, CPU) 501 and one or more memories 502, where at least one instruction is stored in the memories 502, and the at least one instruction is loaded and executed by the processors 501 to implement the following method for classifying titanium alloy stress corrosion damage based on acoustic emission:
s1, acquiring stress corrosion damage signals to be classified by utilizing an acoustic emission sensor.
S2, inputting stress corrosion damage signals into the constructed classification model.
S3, according to the stress corrosion damage signals and the classification model, obtaining classification results of the stress corrosion damage signals, and according to the classification results, obtaining the reasons of the stress corrosion damage of the titanium alloy.
In an exemplary embodiment, a computer readable storage medium, such as a memory comprising instructions executable by a processor in a terminal to perform the above-described acoustic emission based titanium alloy stress corrosion damage classification method is also provided. For example, the computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (10)
1. A method for classifying titanium alloy stress corrosion damage based on acoustic emission, the method comprising:
s1, acquiring stress corrosion damage signals to be classified by using an acoustic emission sensor;
s2, inputting the stress corrosion damage signals into the constructed classification model;
s3, according to the stress corrosion damage signals and the classification model, a classification result of the stress corrosion damage signals is obtained, and according to the classification result, the stress corrosion damage reason of the titanium alloy is obtained.
2. The method according to claim 1, wherein the process of constructing the classification model in S2 includes:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set;
s22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data;
s23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and a decision tree algorithm.
3. The method according to claim 2, wherein the acoustic emission-loaded titanium alloy stress corrosion test in S21 comprises:
s211, fixing a stress corrosion sample in a clamping block on a base of a material deformation in-situ characterization system;
s212, fixing an acoustic emission sensor serving as an experimental probe on the surface of the stress corrosion sample;
and S213, uniformly coating Vaseline coupling agent on the contact part of the experimental probe and the stress corrosion sample, and carrying out a titanium alloy stress corrosion experiment carrying acoustic emission.
4. The method of claim 2, wherein the acoustic emission sensor comprises: broadband sensor, preamplifier and acquisition device PCI-2.
5. The method according to claim 2, wherein the extracting the waveform signal data set according to the experimental data set in S22 includes:
s221, extracting waveform signal data of voltage versus time at each impact from the experimental data set;
s222, preprocessing the waveform signal data;
s223, extracting preset quantity of waveform signal data from the preprocessed waveform signal data, carrying out waveform drawing on the extracted waveform signal data to obtain drawing results, and labeling the waveform signal data according to the drawing results.
6. The method of claim 5, wherein preprocessing the waveform signal data in S222 comprises:
acquiring test time in experimental information in the titanium alloy stress corrosion experimental process, and denoising and filtering the waveform signal data;
and sorting the processed waveform signal data by using codes, naming the file name of the sorted waveform signal data as test time, and obtaining the file content of the waveform signal data from which the experimental information is removed.
7. The method of claim 5, wherein the tag in S223 comprises: burst signals, continuous signals, and noise.
8. The method according to claim 2, wherein the parameter setting of the decision tree algorithm in S23 comprises:
the division standard Criterion of the decision tree is set as a base coefficient gini;
setting the maximum depth of the decision tree as None;
the minimum number of samples on the leaf node of the decision tree is set to 1.
9. A titanium alloy stress corrosion damage classification device based on acoustic emission, the device comprising:
the acquisition module is used for acquiring stress corrosion damage signals to be classified by utilizing the acoustic emission sensor;
the input module is used for inputting the stress corrosion damage signals into the constructed classification model;
and the output module is used for obtaining a classification result of the stress corrosion damage signal according to the stress corrosion damage signal and the classification model and obtaining the cause of the stress corrosion damage of the titanium alloy according to the classification result.
10. The apparatus of claim 9, wherein the input module is configured to:
s21, carrying out a titanium alloy stress corrosion experiment carrying an acoustic emission sensor, and obtaining an experiment data set;
s22, extracting to obtain a waveform signal data set according to the experimental data set; the waveform signal data set comprises marked waveform signal data and unmarked waveform signal data;
s23, obtaining a constructed classification model according to the marked waveform signal data, the unmarked waveform signal data and a decision tree algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311035526.9A CN117191956A (en) | 2023-08-16 | 2023-08-16 | Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311035526.9A CN117191956A (en) | 2023-08-16 | 2023-08-16 | Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117191956A true CN117191956A (en) | 2023-12-08 |
Family
ID=88987830
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311035526.9A Pending CN117191956A (en) | 2023-08-16 | 2023-08-16 | Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117191956A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117647587A (en) * | 2024-01-30 | 2024-03-05 | 浙江大学海南研究院 | Acoustic emission signal classification method, computer equipment and medium |
-
2023
- 2023-08-16 CN CN202311035526.9A patent/CN117191956A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117647587A (en) * | 2024-01-30 | 2024-03-05 | 浙江大学海南研究院 | Acoustic emission signal classification method, computer equipment and medium |
| CN117647587B (en) * | 2024-01-30 | 2024-04-09 | 浙江大学海南研究院 | Acoustic emission signal classification method, computer equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112858473B (en) | Turnout switch blade damage state monitoring method based on feature fusion | |
| Doan et al. | An unsupervised pattern recognition approach for AE data originating from fatigue tests on polymer–composite materials | |
| Wang et al. | Densely connected convolutional networks for vibration based structural damage identification | |
| CN117191956A (en) | Acoustic emission-based titanium alloy stress corrosion damage classification method and apparatus | |
| CN111177655B (en) | Data processing method and device and electronic equipment | |
| Rummel | Nondestructive inspection reliability history, status and future path | |
| Paulraj et al. | Structural steel plate damage detection using non destructive testing, frame energy based statistical features and artificial neural networks | |
| CN110990978A (en) | Bolt state monitoring method and device | |
| CN110797032B (en) | Voiceprint database establishing method and voiceprint identification method | |
| CN114002332A (en) | Structural damage monitoring and early warning method and structural integrity digital twinning system | |
| CN117347501A (en) | Building material performance detection system and method | |
| CN116881712A (en) | Electromagnetic pulse signal identification method for movable cracks of concrete dam | |
| Nunes et al. | Acoustic structural integrity assessment of ceramics using supervised machine learning and uncertainty-based rejection | |
| Ma et al. | A percussion method with attention mechanism and feature aggregation for detecting internal cavities in timber | |
| Barat et al. | Discovering data flow discords for enhancing noise immunity of acoustic-emission testing | |
| Shahsavari et al. | Structural health monitoring of a vertical lift bridge using vibration data | |
| CN114373452A (en) | Voice abnormity identification and evaluation method and system based on deep learning | |
| CN117233347B (en) | Carbon steel spheroidization grade measuring method, system and equipment | |
| CN114383834A (en) | Ocean engineering structure micro-damage judgment method | |
| Sharma et al. | Acoustic Sensor-Based Approach for Detecting Damage in Masonry Structures | |
| CN117647587B (en) | Acoustic emission signal classification method, computer equipment and medium | |
| CN115144259B (en) | Method and system for detecting deformation resistance of steel | |
| CN114910553A (en) | Memory, and identification method, device and equipment for hydrogen induced cracking process | |
| Chen et al. | Acoustic emission signal-based non-destructive testing of carbon content of Pr-Nd alloys | |
| JP6429215B1 (en) | Inspection method for structures |
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 |