Equipment wall damage identification methods, devices, electronic equipment and readable storage media

Abstract

This invention provides a method, apparatus, electronic device, and readable storage medium for identifying equipment wall damage, belonging to the field of acoustic monitoring technology. The method includes: acquiring acoustic signals of the equipment wall over a period of time; obtaining entropy parameters and acoustic reconstruction data based on the acoustic signals; determining a first theoretical damage state of the equipment wall based on the time domain, frequency domain, and entropy parameters of the acoustic signals; determining a second theoretical damage state of the equipment wall based on the acoustic signals and acoustic reconstruction data; and determining the overall damage state of the equipment wall based on the first and second theoretical damage states, where the damage state includes both the presence and absence of damage. This invention enables effective identification of equipment wall defect propagation signals under complex environments and strong background noise, achieving multi-dimensional acoustic identification of damage states, improving the accuracy of acoustic signal identification in strong background noise environments, effectively improving the accuracy of acoustic monitoring, and providing a basis for equipment operation and maintenance planning.

Description

Technical Field

[0001] This invention relates to the field of acoustic monitoring technology, specifically to a method for identifying damage to the wall of an equipment, a device for identifying damage to the wall of an equipment, an electronic device, and a readable storage medium. Background Technology

[0002] During the manufacturing process, the equipment wall is difficult to machine, making it prone to hidden defects. During operation, the combined effects of high temperature, high pressure, and various media can inevitably lead to damage and cracks, making the operation of the equipment with defects a significant potential risk.

[0003] Cracks in the vessel wall are random and initially very small, exceeding the detection limit of flaw detection methods and thus impossible to detect accurately. Furthermore, these cracks are often concealed, appearing on the inner and outer surfaces of equipment and pipelines, and even inside them. However, high-temperature equipment is typically covered with insulation, making routine inspections difficult. Because the initial cracks are small and difficult to detect, and their propagation rate is rapid in the later stages, if not detected and controlled in time, they can easily lead to major accidents such as equipment penetration, leakage, or even fire and explosion within a short period.

[0004] Defective equipment operation is unavoidable, and active defects pose a high risk. Early detection of damage, monitoring of the damage process, and timely intervention are crucial for safe production. Online monitoring technology is of paramount importance. Acoustic emission (AE) determines the material's condition by detecting transient elastic waves emitted due to rapid energy release in localized areas of the material. Array sensors can be used for overall monitoring. It is applied to online monitoring of atmospheric pressure storage tanks and hydraulic testing of pressure vessels. However, due to significant interference from ambient noise and other factors at the equipment site, effective extraction and identification of damage signals are difficult, making AE technology unsuitable for online monitoring of equipment damage conditions, or resulting in inaccurate monitoring during application. Summary of the Invention

[0005] The purpose of this invention is to provide a method, apparatus, electronic device, and readable storage medium for identifying equipment wall damage, so as to at least solve the problems mentioned above, which are limited by the large interference from noise and other factors at the device site, making it difficult to effectively extract and identify damage and defect signals, thus making it difficult to apply acoustic emission technology to online monitoring of equipment damage status, or making the monitoring inaccurate when used.

[0006] To achieve the above objectives, a first aspect of the present invention provides a method for identifying damage to the wall of a device, the method comprising: Acquire the acoustic signal of the device wall over a certain operating time; Based on the acoustic signal, the entropy parameter and acoustic reconstruction data of the acoustic signal are obtained; Based on the time domain, frequency domain, and entropy parameter of the acoustic signal, the first theoretical damage state of the device wall is determined; Based on the acoustic signal and the acoustic reconstruction data, the second theoretical damage state of the vessel wall is determined; Based on the first theoretical damage state and the second theoretical damage state, the damage state of the vessel wall is determined, including the presence of damage and the absence of damage.

[0007] Optionally, based on the acoustic signal, the entropy parameter of the acoustic signal is obtained, including: The entropy parameter includes the signal entropy, which is calculated using the following formula:

[0008] in, For signal entropy; Let be the probability of an acoustic signal occurring with respect to the i-th voltage value of an acoustic emission signal.

[0009] Optionally, based on the acoustic signal, acoustic reconstruction data of the acoustic signal is obtained, including: The acoustic signal is used as input to the data reconstruction model to obtain acoustic reconstruction data of the acoustic signal; The data reconstruction model is obtained by training the autoencoder based on a training dataset, which includes time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture.

[0010] Optionally, based on the time domain, frequency domain, and entropy parameter of the acoustic signal, a first theoretical damage state of the vessel wall is determined, including: Using the time-domain, frequency-domain, and entropy parameters of the acoustic signal as inputs to the hypersphere model, the generalized distance between the acoustic signal and the center of the hypersphere model is obtained. If the generalized distance is less than or equal to the generalized radius of the hypersphere of the hypersphere model, then the first theoretical damage state of the device wall is determined to be no damage. If the generalized distance is greater than the generalized radius of the hypersphere in the hypersphere model, then the first theoretical damage state of the device wall is determined to be that damage exists.

[0011] Optionally, the hypersphere model is trained using the following method: Construct a training dataset, which includes the time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture. The hypersphere model is obtained by training SVDD on the training dataset.

[0012] Optionally, the expression for the hypersphere model is:

[0013] Where, d The generalized distance between the acoustic signal and the center of the hypersphere of the hypersphere model is given. and For Lagrange multipliers; K For kernel functions; It is a constant; z For the feature values ​​of the test data, and These are the eigenvalues ​​of the time-domain, frequency-domain, and entropy parameters of the acoustic signal.

[0014] Optionally, based on the acoustic signal and the acoustic reconstruction data, a second theoretical damage state of the vessel wall is determined, including: Calculate the mean square error between the acoustic signal and the acoustic reconstruction data; If the mean square error is less than or equal to the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be no damage. If the mean square error is greater than the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be that damage exists.

[0015] Optionally, based on the first theoretical damage state and the second theoretical damage state, the damage state of the vessel wall is determined, including: If both the first theoretical damage state and the second theoretical damage state are present, then the damage state of the vessel wall is determined to be present. If both the first theoretical damage state and the second theoretical damage state are non-damaged, then the damage state of the vessel wall is determined to be non-damaged. If the first theoretical damage state and the second theoretical damage state are respectively the presence of damage and the absence of damage, then the damage state of the vessel wall is determined based on the generalized distance and the root mean square error.

[0016] Optionally, based on generalized distance and mean square error, the damage state of the vessel wall can be determined, including: The judgment coefficient is calculated based on the generalized distance and the mean square error; If the judgment coefficient is less than the preset threshold, the damage state of the device wall is determined to be no damage. If the judgment coefficient is greater than or equal to the preset threshold, then the damage state of the device wall is determined to be present.

[0017] Optionally, the judgment coefficient can be calculated using the following formula: ; in, y The judgment coefficient; d For generalized distance; eThe mean squared error; t To reconstruct the threshold.

[0018] Optionally, if the damage status of the equipment wall is determined to be present: The damage stress is determined based on the total acoustic emission count of the acoustic signal; the degree of damage to the device wall is determined based on the damage stress.

[0019] If the damage stress is less than or equal to the first preset stress, then the degree of damage to the equipment wall during this operating period is determined to be initial damage. If the damage stress is greater than the first preset stress and less than or equal to the second preset stress, then the degree of damage to the equipment wall during this operating period is determined to be low damage. If the damage stress is greater than the second preset stress and less than or equal to the third preset stress, then the degree of damage to the equipment wall during this operating period is determined to be medium damage. If the damage stress is greater than the third preset stress, then the damage level of the equipment wall during this operating period is determined to be high damage.

[0020] Optionally, the damage stress can be calculated using the following formula:

[0021] in, Damage stress; This represents the total acoustic emission count; D and C are constants. These are the fitting coefficients; The material strength of the equipment wall; This is the temperature influence coefficient; This is the pressure influence coefficient; This represents the flow rate impact coefficient.

[0022] A second aspect of the present invention provides a device for identifying damage to the device wall, the device comprising: The parameter acquisition module is used to acquire the acoustic signal of the device wall during a certain operating time. The data determination module is used to obtain the entropy parameter and acoustic reconstruction data of the acoustic signal based on the acoustic signal; The first theoretical damage state determination module is used to determine the first theoretical damage state of the device wall based on the time domain, frequency domain and entropy parameter of the acoustic signal; The second theoretical damage state determination module is used to determine the second theoretical damage state of the device wall based on the acoustic signal and the acoustic reconstruction data. The damage state determination module is used to determine the damage state of the vessel wall based on the first theoretical damage state and the second theoretical damage state, wherein the damage state of the vessel wall includes the presence of damage and the absence of damage.

[0023] A third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described device wall damage identification method.

[0024] On the other hand, the present invention provides a readable storage medium storing instructions for causing a machine to perform the above-described device wall damage identification method.

[0025] This technical solution determines the damage state of the vessel wall based on the time domain, frequency domain, entropy parameters, and acoustic reconstruction data of the acoustic signal. It enables effective identification of the extended signal of the vessel wall defect under complex environments and strong background noise, achieves multi-dimensional acoustic identification of the damage state, improves the accuracy of acoustic signal identification under strong background noise environment, and further improves the accuracy of acoustic monitoring, providing a basis for equipment operation and maintenance planning.

[0026] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0027] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the device wall damage identification method provided by the present invention; Figure 2 This is a flowchart of the device wall damage identification method provided by the present invention; Figure 3 This is a schematic diagram of the device wall damage identification device provided by the present invention; Figure 4 This is a schematic diagram illustrating the changes in different acoustic signals when the unfused entropy parameter is provided by the present invention; Figure 5 This is a schematic diagram illustrating the changes in different acoustic signals after fusing the entropy parameter provided by the present invention.

[0028] Explanation of reference numerals in the attached figures 10 - Parameter acquisition module; 20 - Data determination module; 30 - First theoretical damage state determination module; 40 - Second theoretical damage state determination module; 50 - Damage state determination module. Detailed Implementation

[0029] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0030] Figure 1 This is a flowchart of the device wall damage identification method provided by the present invention; Figure 2 This is a flowchart of the device wall damage identification method provided by the present invention; Figure 3 This is a schematic diagram of the device wall damage identification device provided by the present invention; Figure 4 This is a schematic diagram illustrating the changes in different acoustic signals when the unfused entropy parameter is provided by the present invention; Figure 5 This is a schematic diagram illustrating the changes in different acoustic signals after fusing the entropy parameter provided by the present invention.

[0031] Example 1 like Figure 1-2 As shown, an embodiment of the present invention provides a method for identifying damage to the wall of a device, the method comprising: Step 1: Acquire the acoustic signal of the device wall over a certain operating time; Step 2: Based on the acoustic signal, obtain the entropy parameter and acoustic reconstruction data of the acoustic signal; Step 3: Based on the time domain, frequency domain, and entropy parameter of the acoustic signal, determine the first theoretical damage state of the device wall. Step 4: Based on the acoustic signal and the acoustic reconstruction data, determine the second theoretical damage state of the vessel wall; Step 5: Based on the first theoretical damage state and the second theoretical damage state, determine the damage state of the vessel wall, which includes the presence of damage and the absence of damage.

[0032] Specifically, in the prior art, acoustic signals can be acquired by acoustic sensors, which can be set close to the container wall or directly on the container wall to ensure that the acquired acoustic signals are more accurate.

[0033] Specifically, in existing technologies, cracks in vessel walls are random and initially very small, exceeding the detection limit of flaw detection methods and thus impossible to detect accurately. Furthermore, these cracks are often concealed, potentially appearing on the inner and outer surfaces of equipment and pipelines, or even inside them. However, high-temperature equipment is typically covered with insulation, making routine inspections difficult. Because cracks are initially small and difficult to detect, and their propagation rate is rapid in the later stages, failure to detect them and take timely measures to control process operations can easily lead to major accidents such as equipment penetration, leakage, or even fires and explosions within a short period.

[0034] Damage assessment of equipment walls using acoustic signals typically relies solely on time-domain parameters or frequency-domain filtering. However, during equipment operation, interference from nonlinear factors such as ambient noise and operating noise makes it difficult to accurately extract the damage acoustic signals, leading to errors in the final assessment. Therefore, this embodiment obtains the corresponding entropy parameter from the acoustic signal, combines the time and frequency domains of the acoustic signal with the corresponding acoustic reconstruction data for comprehensive assessment, effectively extracting damage characteristic signals from the data to accurately determine the damage state of the equipment wall, thereby identifying whether the wall is in a damaged state or not.

[0035] like Figure 4-5 As shown, although the acoustic emission signals from damage to the vessel wall and noise signals such as disturbance interference, electromagnetic interference, and material friction have certain differences, their time-domain and frequency-domain characteristics overlap significantly, making them difficult to distinguish. Furthermore, it was found that the damage signals exhibit randomness and complexity, while background noise and electromagnetic interference signals often show periodicity. Entropy can describe the complexity of information; therefore, we established an entropy parameter characterization model. After adding the entropy parameter, the signal interface becomes clearer, improving the time-frequency spatial identification of the damage acoustic signals. The damage signal distribution is relatively concentrated, while the noise signal distribution is wider. The damage entropy parameter value is significantly lower than that of the noise signal, enabling accurate identification of the damage signal.

[0036] More specifically, in this embodiment, taking the wall of a certain device as an example, 1500 experimental data points of welding cold cracks were collected from the device. These data included 1500 electromagnetic interference signals and 1500 bubble burst noise signals. By judging solely through the time and frequency domains of the aforementioned data, 696 crack propagation data points were successfully identified, achieving an accuracy rate of 46.4%. Using the combined time, frequency, and entropy parameter method, 1470 crack propagation data points were successfully identified, achieving an accuracy rate of 98%.

[0037] Further, in this embodiment, based on the acoustic signal, the entropy parameter of the acoustic signal is obtained, including: The entropy parameter includes the signal entropy, which is calculated using the following formula:

[0038] in, For signal entropy; Let be the probability of an acoustic signal occurring with respect to the i-th voltage value of an acoustic emission signal.

[0039] Preferably, the signal entropy can also be replaced by conditional entropy, cross entropy, relative entropy, etc.

[0040] Specifically, in this embodiment, the entropy parameter of the acoustic signal is extracted by establishing a multimodal damage acoustic signal feature extraction method based on time-frequency entropy.

[0041] The acoustic reconstruction data of the acoustic signal, obtained based on the acoustic signal, includes: The acoustic signal is used as input to the data reconstruction model to obtain acoustic reconstruction data of the acoustic signal; The data reconstruction model is obtained by training the autoencoder based on a training dataset, which includes time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture.

[0042] Further, in the above steps, based on the time domain, frequency domain, and entropy parameter of the acoustic signal, the first theoretical damage state of the device wall is determined, including: Step 31: Use the time domain, frequency domain, and entropy parameters of the acoustic signal as input to the hypersphere model to obtain the generalized distance between the acoustic signal and the center of the hypersphere model. Step 32: Based on the generalized distance, determine the damage state of the vessel wall, which includes the presence of damage and the absence of damage.

[0043] Further, in step 32, based on the generalized distance, the damage state of the vessel wall is determined, including the presence of damage and the absence of damage, including: If the generalized distance is less than or equal to the generalized radius of the hypersphere of the hypersphere model, then the damage state of the device wall is determined to be no damage. If the generalized distance is greater than the generalized radius of the hypersphere in the hypersphere model, then the damage state of the device wall is determined to be that damage exists.

[0044] Specifically, when the generalized distance between the acoustic signal and the center of the hypersphere of the hypersphere model is less than or equal to the generalized radius of the hypersphere, it indicates that the data is inside the hypersphere, and the online monitoring data is judged to be normal data. At this time, the device wall damage state is no damage state. When the generalized distance between the acoustic signal and the center of the hypersphere of the hypersphere model is greater than the generalized radius of the hypersphere, it indicates that the data is outside the hypersphere, and the online monitoring data is judged to be abnormal data. At this time, the device wall damage state is present.

[0045] Furthermore, the hypersphere model is trained using the following method: Construct a training dataset, which includes the time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture. The hypersphere model is obtained by training SVDD on the training dataset.

[0046] The entropy parameter obtained from historical acoustic signals is the same as described above, i.e., the conversion is performed using the formulas mentioned above, including:

[0047] in, It is the entropy parameter; Let be the probability of an acoustic signal occurring with respect to the i-th voltage value of an acoustic emission signal.

[0048] Specifically, to improve model accuracy, a database of acoustic emission monitoring characteristic signals for damage dynamic processes, adapted for incremental training, was established, comprising 1200 data files and over 100,000 data entries. This database covers primary test data for acoustic emission monitoring of typical damage in typical materials, including tensile, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture; typical damage sample data; extended characteristic data of typical damage samples; abnormal noise data; and test sample data.

[0049] To address the challenges of high-dimensional data processing and accurate mapping of monitoring signals from three-dimensional sensor arrays in high-parameter pressure vessels, a proposed method is to improve the Support Vector Machine (SVM) with an adaptive Gaussian kernel function bandwidth to establish an SVDD support vector description model. An SVDD hypersphere of suitable volume is trained, and key information from time-frequency entropy multimodal feature vectors is extracted to form a mapping relationship between the high-dimensional feature space matrix and the hypersphere description boundary.

[0050] More specifically, an SVDD hypersphere of suitable volume is trained to form a mapping relationship between the high-dimensional feature space matrix and the hypersphere's boundary description. To ensure a close fit between the hypersphere and the data distribution, Gaussian kernel function optimization is used to map the original data to a high-dimensional space. Then, a hypersphere is found in the high-dimensional space using the method described above, and finally, the hypersphere is inversely mapped back to the original data space to obtain the contour.

[0051] By optimizing the calculation and substituting the kernel function, we obtain the formula for the distance from a point in the original data to the center of the hypersphere after mapping it to a high-dimensional space. This forms the mapping relationship between the high-dimensional feature space matrix and the hypersphere description boundary.

[0052] Further, a sliding composite moving window is used to calculate the real-time trend factor and correct the hypersphere boundary.

[0053] Before each sliding of the composite moving window, the value of the adaptive regularization parameter λ is determined; then, L1TF is applied to the data within the composite moving window, and the absolute value of the derivative of the filtered line segment is defined as the trend factor (TF). The last point within the composite moving window is an occasional anomaly. If there is only one anomaly, it is unlikely to affect the overall trend of all 50 points. The absolute value of the slope of the filtered line segment, |K1|, is relatively small, thus reducing the impact of occasional anomalies. Generally, when an equipment experiences an early failure, its characteristic indicators will gradually deviate from the normal operating state in the same direction. When five consecutive data points deviate from the normal operating state, although the deviation of a single point is not large, the absolute value of the slope of the filtered line segment, |K2|, is relatively large, thus achieving the purpose of correcting the hypersphere and enabling a more accurate distinction between damage signals such as crack propagation and interference signals.

[0054] SVDD is a classic single classifier. Its main idea is as follows: First, the low-dimensional training data (not less than one dimension) is mapped to a high-dimensional space using a kernel function. Then, a hypersphere with center *a* and radius *r* is found in the high-dimensional space, such that most of the positive samples (feature values ​​of historical normal operating vibration data) in the training dataset are contained within the hypersphere, while excluding occasional outlier samples as much as possible. An infinitely large hypersphere would contain all training samples, but it would also include negative samples (feature values ​​of occasional outlier vibration data). Therefore, a balance needs to be struck between the size of the hypersphere and the ability to distinguish occasional outliers. In other words, the hypersphere's descriptive boundary needs to capture all positive samples, while minimizing its volume. Therefore, the above optimization problem can be described using formulas (3-1) and (3-2): (3-1) In the formula, F(•) is the error function. To meet the above requirements, the error function needs to be minimized, thereby including as many positive class samples as possible within the hypersphere. To achieve this goal, the constraint of equation (3-2) needs to be satisfied: (3-2) In the formula, For the training dataset, l It is the number of samples in the training dataset.

[0055] Since occasional outliers inevitably appear in the training dataset, the distance from the training sample to the center of the hypersphere should not strictly satisfy formula (3-2). To address the possibility of outliers, a slack variable is introduced. To penalize it, update formula (3-2) as shown in formula (3-3): (3-3) Introducing slack variables ( After ≥0), some outliers can be allowed to be outside the hypersphere. The error function of formula (3-1) no longer meets the requirements, and it is updated as shown in formula (3-4); a penalty factor C is introduced to control the slack variables. This limits the influence of a small number of negative samples on the size of the hypersphere.

[0056] (3-4) By incorporating formula (3-3) into formula (3-4) using the Lagrange multiplier, the above optimization problem is transformed into an unconstrained optimization problem, and the resulting objective function is shown in formula (3-5): (3-5) In the formula, ≥0, For all operators ≥ 0, a Lagrange multiplier is used to solve the aforementioned optimization problem. L Relative to R, a , Minimize, relative to , Maximize. L For R, a , Setting the partial derivatives to 0 yields the following constraints: (3-6) because ,and ≥0, ≥0, therefore we can derive the Lagrange multiplier operator. and C The size relationship is shown in formula (3-7): (3-7) Substituting equation (3-6) into equation (3-4) yields the optimized function: (3-8) In the formula, It is a kernel function. Kernel functions can map the training dataset from a low-dimensional space to a high-dimensional space, allowing the training data to find a hypersphere of suitable size in the high-dimensional space. This paper uses the Gaussian kernel function as the SVDD kernel function, and its expression is as follows: (3-9) As shown in formula (3-10), when the training samples are inside the hypersphere, = 0; when the training samples are on the description boundary of the hypersphere, 0 < <C, called SV(on); when the training sample is outside the hypersphere description boundary, = C, called SV(out). Samples on and outside the hypersphere interface are all support vectors (SV'S): (3-10) Furthermore, the expression for the hypersphere model is:

[0057] in, d The generalized distance between the acoustic signal and the center of the hypersphere of the hypersphere model is given. and For Lagrange multipliers; K For kernel functions; It is a constant; z For the feature values ​​of the test data, and These are the eigenvalues ​​of the time-domain, frequency-domain, and entropy parameters of the acoustic signal.

[0058] Furthermore, the generalized radius of a hypersphere r The distance from any support vector on the hypersphere's boundary to the hypersphere's center 'a' can be calculated. Therefore, the generalized radius of the hypersphere in the hypersphere model is:

[0059] in, d r Let be the generalized radius of the acoustic signal relative to the center of the hypersphere in the hypersphere model; and For Lagrange multipliers; K For kernel functions; It is a constant; These are support vectors; and These are the eigenvalues ​​of the time-domain, frequency-domain, and entropy parameters of the acoustic signal.

[0060] In step four, based on the acoustic signal and the acoustic reconstruction data, the second theoretical damage state of the vessel wall is determined, including: Calculate the mean square error between the acoustic signal and the acoustic reconstruction data; If the mean square error is less than or equal to the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be no damage. If the mean square error is greater than the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be that damage exists.

[0061] Finally, if both the first theoretical damage state and the second theoretical damage state are present, then the damage state of the vessel wall is determined to be present. If both the first theoretical damage state and the second theoretical damage state are non-damaged, then the damage state of the vessel wall is determined to be non-damaged. If the first theoretical damage state and the second theoretical damage state are respectively the presence of damage and the absence of damage, then the damage state of the vessel wall is determined based on the generalized distance and the root mean square error.

[0062] Furthermore, based on generalized distance and mean square error, the damage state of the vessel wall is determined, including: The judgment coefficient is calculated based on the generalized distance and the mean square error; If the judgment coefficient is less than the preset threshold, the damage state of the device wall is determined to be no damage. If the judgment coefficient is greater than or equal to the preset threshold, then the damage state of the device wall is determined to be present.

[0063] By using the above method, the corresponding damage state is calculated using two unrelated parameters, and the final judgment result is obtained based on the specific damage state, which can ensure that the damage identification result is more accurate.

[0064] Furthermore, the judgment coefficient is calculated using the following formula: ; in, y The judgment coefficient; d For generalized distance; e The mean squared error; t To reconstruct the threshold.

[0065] Specifically, in this embodiment, the judgment coefficient is calculated using the above-mentioned calculation formula, which can ensure the accuracy of the calculation results and improve the recognition results.

[0066] Example 2 In this embodiment, when the damage state of the equipment wall is determined to be present: The damage stress is determined based on the total acoustic emission count of the acoustic signal; the degree of damage to the device wall is determined based on the damage stress.

[0067] If the damage stress is less than or equal to the first preset stress, then the degree of damage to the equipment wall during this operating period is determined to be initial damage. If the damage stress is greater than the first preset stress and less than or equal to the second preset stress, then the degree of damage to the equipment wall during this operating period is determined to be low damage. If the damage stress is greater than the second preset stress and less than or equal to the third preset stress, then the degree of damage to the equipment wall during this operating period is determined to be medium damage. If the damage stress is greater than the third preset stress, then the damage level of the equipment wall during this operating period is determined to be high damage.

[0068] The damage stress is calculated using the following formula:

[0069] in, Damage stress; This represents the total acoustic emission count; D and C are constants. These are the fitting coefficients; The material strength of the equipment wall; This is the temperature influence coefficient; This is the pressure influence coefficient; This represents the flow rate impact coefficient.

[0070] In this embodiment, when the damage to the equipment wall is classified as initial damage, it indicates the least severe damage, generally considered to be in the crack nucleation stage, at which point the risk is low. When the damage is classified as low damage, it is generally considered to be in the early stage of steady-state propagation, posing a certain degree of risk and requiring continuous monitoring. When the damage is classified as moderate damage, it is generally considered to be in the late stage of steady-state propagation, a stage with higher risk, requiring depressurization and retesting to reduce operational risks. When the damage is classified as medium damage, it is generally considered to be in unstable propagation, a stage with the highest risk, requiring immediate shutdown and maintenance to reduce safety risks. Therefore, based on the degree of equipment wall damage, four alarm levels are established, covering different damage stages and enabling accurate early warning.

[0071] Example 3 like Figure 3 As shown, this embodiment also provides a device for identifying damage to the device wall, the device comprising: The parameter acquisition module 10 is used to acquire the acoustic signal of the device wall during a certain operating time. The data determination module 20 is used to obtain the entropy parameter and acoustic reconstruction data of the acoustic signal based on the acoustic signal; The first theoretical damage state determination module 30 is used to determine the first theoretical damage state of the device wall based on the time domain, frequency domain and entropy parameter of the acoustic signal. The second theoretical damage state determination module 40 is used to determine the second theoretical damage state of the device wall based on the acoustic signal and the acoustic reconstruction data. The damage state determination module 50 is used to determine the damage state of the vessel wall based on the first theoretical damage state and the second theoretical damage state, wherein the damage state of the vessel wall includes the presence of damage and the absence of damage.

[0072] Example 4 This embodiment also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described device wall damage identification method.

[0073] Example 5 This embodiment also provides a readable storage medium storing instructions for causing a machine to perform the above-described device wall damage identification method.

[0074] Example 6 In this embodiment, the same data was used to verify the effect. The accuracy rate of analysis using conventional neural network models directly using the collected acoustic signals was 83%, while the recognition accuracy using the technical means of this solution reached 98%, which is an effective improvement.

[0075] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0076] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application.

[0077] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0078] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for identifying damage to the wall of equipment, characterized in that, The method includes: Acquire the acoustic signal of the device wall over a certain operating time; Based on the acoustic signal, the entropy parameter and acoustic reconstruction data of the acoustic signal are obtained; Based on the time domain, frequency domain, and entropy parameter of the acoustic signal, the first theoretical damage state of the device wall is determined; Based on the acoustic signal and the acoustic reconstruction data, the second theoretical damage state of the vessel wall is determined; Based on the first theoretical damage state and the second theoretical damage state, the damage state of the vessel wall is determined, including the presence of damage and the absence of damage.

2. The equipment wall damage identification method according to claim 1, characterized in that, Based on the acoustic signal, the entropy parameter of the acoustic signal is obtained, including: The entropy parameter includes the signal entropy, which is calculated using the following formula: in, The signal entropy; Let be the probability of an acoustic signal occurring with respect to the i-th voltage value of an acoustic emission signal.

3. The equipment wall damage identification method according to claim 1, characterized in that, Based on the acoustic signal, acoustic reconstruction data of the acoustic signal is obtained, including: The acoustic signal is used as input to the data reconstruction model to obtain acoustic reconstruction data of the acoustic signal; The data reconstruction model is obtained by training the autoencoder based on a training dataset, which includes time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture.

4. The equipment wall damage identification method according to claim 1, characterized in that, Based on the time domain, frequency domain, and entropy parameter of the acoustic signal, the first theoretical damage state of the vessel wall is determined, including: Using the time-domain, frequency-domain, and entropy parameters of the acoustic signal as inputs to the hypersphere model, the generalized distance between the acoustic signal and the center of the hypersphere model is obtained. If the generalized distance is less than or equal to the generalized radius of the hypersphere of the hypersphere model, then the first theoretical damage state of the device wall is determined to be no damage. If the generalized distance is greater than the generalized radius of the hypersphere in the hypersphere model, then the first theoretical damage state of the device wall is determined to be that damage exists.

5. The equipment wall damage identification method according to claim 4, characterized in that, The hypersphere model was trained using the following method: Construct a training dataset, which includes the time-domain, frequency-domain, and entropy parameters of historical acoustic signals generated by material tension, three-point bending, stress corrosion, flow erosion, hydrogen-induced cracking, and fatigue fracture. The hypersphere model is obtained by training SVDD on the training dataset.

6. The equipment wall damage identification method according to claim 4, characterized in that, The expression for the hypersphere model is: Where, d The generalized distance between the acoustic signal and the center of the hypersphere of the hypersphere model is given. and For Lagrange multipliers; K For kernel functions; It is a constant; z For the feature values ​​of the test data, and These are the eigenvalues ​​of the time-domain, frequency-domain, and entropy parameters of the acoustic signal.

7. The equipment wall damage identification method according to claim 4, characterized in that, Based on the acoustic signal and the acoustic reconstruction data, the second theoretical damage state of the vessel wall is determined, including: Calculate the mean square error between the acoustic signal and the acoustic reconstruction data; If the mean square error is less than or equal to the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be no damage. If the mean square error is greater than the reconstruction threshold, then the second theoretical damage state of the device wall is determined to be that damage exists.

8. The equipment wall damage identification method according to claim 7, characterized in that, Based on the first theoretical damage state and the second theoretical damage state, the damage state of the vessel wall is determined, including: If both the first theoretical damage state and the second theoretical damage state are present, then the damage state of the vessel wall is determined to be present. If both the first theoretical damage state and the second theoretical damage state are non-damaged, then the damage state of the vessel wall is determined to be non-damaged. If the first theoretical damage state and the second theoretical damage state are respectively the presence of damage and the absence of damage, then the damage state of the vessel wall is determined based on the generalized distance and the root mean square error.

9. The equipment wall damage identification method according to claim 8, characterized in that, Based on generalized distance and mean square error, the damage state of the vessel wall is determined, including: The judgment coefficient is calculated based on the generalized distance and the mean square error; If the judgment coefficient is less than the preset threshold, the damage state of the device wall is determined to be no damage. If the judgment coefficient is greater than or equal to the preset threshold, then the damage state of the device wall is determined to be present.

10. The equipment wall damage identification method according to claim 9, characterized in that, The judgment coefficient is calculated using the following formula: ； in, y The judgment coefficient; d For generalized distance; e The mean squared error; t This is for reconstructing the threshold.

11. The equipment wall damage identification method according to claim 1, characterized in that, The method further includes: when the damage state of the equipment wall is determined to be present: The damage stress is determined based on the total acoustic emission count of the acoustic signal; the degree of damage to the device wall is determined based on the damage stress. If the damage stress is less than or equal to the first preset stress, then the degree of damage to the equipment wall during this operating period is determined to be initial damage. If the damage stress is greater than the first preset stress and less than or equal to the second preset stress, then the degree of damage to the equipment wall during this operating period is determined to be low damage. If the damage stress is greater than the second preset stress and less than or equal to the third preset stress, then the degree of damage to the equipment wall during this operating period is determined to be medium damage. If the damage stress is greater than the third preset stress, then the damage level of the equipment wall during this operating period is determined to be high damage.

12. The equipment wall damage identification method according to claim 11, characterized in that, The damage stress is calculated using the following formula: in, Damage stress; This represents the total acoustic emission count; D and C are constants. These are the fitting coefficients; The material strength of the equipment wall; This is the temperature influence coefficient; This is the pressure influence coefficient; This represents the flow rate impact coefficient.

13. A device for identifying damage to the wall of an equipment, characterized in that, The device includes: The parameter acquisition module is used to acquire the acoustic signal of the device wall during a certain operating time. The data determination module is used to obtain the entropy parameter and acoustic reconstruction data of the acoustic signal based on the acoustic signal; The first theoretical damage state determination module is used to determine the first theoretical damage state of the device wall based on the time domain, frequency domain and entropy parameter of the acoustic signal; The second theoretical damage state determination module is used to determine the second theoretical damage state of the device wall based on the acoustic signal and the acoustic reconstruction data. The damage state determination module is used to determine the damage state of the vessel wall based on the first theoretical damage state and the second theoretical damage state, wherein the damage state of the vessel wall includes the presence of damage and the absence of damage.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the device wall damage identification method according to any one of claims 1-12.

15. A readable storage medium storing instructions for causing a machine to perform the device wall damage identification method according to any one of claims 1-12.

Images