Flange stress analysis method and system
Through edge detection, finite element model and acoustic emission signal processing, the accuracy of flange stress analysis is improved, the problem of insufficient analysis in the prior art is solved, and more accurate stress analysis results are achieved.
Patent Information
- Application Number
- CN202411124499.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-15
AI Technical Summary
The existing flange stress analysis methods lack practical foundations, resulting in insufficient analysis accuracy and different idealized results from actual data.
By collecting flange images for edge detection and image segmentation, a finite element model is constructed for force analysis, identify the position-force-affected relationship between the detection edge and the force cloud map, calculate the connotation modal component of the acoustic emission signal and extract the inverted Spectral parameters, input the stress analysis model to output the stress analysis value of the stress monitoring point.
On the basis of actual conditions, the accuracy of flange stress analysis is improved, noise signals are eliminated, key signals are retained, and the neural network model is used to intelligently output the stress analysis results.
Smart Images

Figure CN119026418A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of flange stress analysis, and in particular to a flange stress analysis method and system. Background Art
[0002] The process of flange stress analysis is to analyze the stress applied by the equipment to the flange when the flange is installed on other equipment, so as to warn that the flange cannot withstand the external force.
[0003] At present, the process of stress analysis of flanges is to make a three-dimensional model of the flange, convert irregular surfaces and lines in the three-dimensional model into regular surfaces and lines, and use mathematical formulas to calculate the magnitude and direction of the forces on these surfaces and lines. In this process, only the three-dimensional model of the flange and the process of converting irregular surfaces and lines in the three-dimensional model into regular surfaces and lines can be realized by intelligent machines, while the analysis of the forces requires artificial brain calculations. Secondly, the process of using mathematical formulas to calculate the magnitude and direction of the forces on these surfaces and lines is realized on the basis of the three-dimensional model, that is, the process of constructing a force diagram. The force results obtained through this process are often idealized force results, and there will be discrepancies between the actual force data and the idealized force data. Therefore, due to the lack of practical basis for the stress analysis of flanges, the accuracy of the stress analysis of flanges is insufficient. Summary of the invention
[0004] In order to solve the above problems, the present invention provides a flange stress analysis method and system, which can improve the accuracy of flange stress analysis on a practical basis.
[0005] In a first aspect, the present invention provides a flange stress analysis method, comprising:
[0006] Acquire a flange image of the flange, perform edge detection on the flange image to obtain a detected edge of the flange image, and perform image segmentation on the flange image using the detected edge to obtain a segmented image;
[0007] Identify the segmented area of the flange from the segmented image, construct a finite element model of the segmented area, perform force analysis on the finite element model to obtain a standard force value of the segmented area, and construct a force cloud map of the segmented area based on the standard force value;
[0008] Identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship;
[0009] Collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component;
[0010] The cepstrum parameters are input into a stress analysis model, so as to use the stress analysis model to output stress analysis values of the stress monitoring points, and the stress analysis values are used as stress analysis results of the flange.
[0011] In a possible implementation manner of the first aspect, performing edge detection on the flange image to obtain a detected edge of the flange image includes:
[0012] The edge value of the flange image is calculated using the following formula:
[0013]
[0014] Wherein, G represents the edge value, G x represents the gradient value of filtering the flange image I in the horizontal direction, G y represents the gradient value of filtering the flange image I in the vertical direction;
[0015] When the edge value is greater than a preset edge value, a pixel point corresponding to the edge value in the flange image is used as an edge pixel point;
[0016] Calculating the pixel similarity between each edge pixel in the edge pixels and the neighboring pixels of the edge pixel;
[0017] Based on the pixel similarity, each edge pixel in the edge pixel is connected pixel by pixel to obtain the detected edge.
[0018] In a possible implementation manner of the first aspect, constructing the finite element model of the subdivided area includes:
[0019] Collecting point cloud data of the subdivided area;
[0020] The point cloud coordinates of the point cloud data are mapped to a two-dimensional coordinate system using the following formula to obtain two-dimensional coordinates:
[0021]
[0022] Wherein, (u, v) represents the two-dimensional coordinates, and (x, y, z) represents the point cloud coordinates of the point cloud data;
[0023] Based on the two-dimensional coordinates, repair the point cloud data to obtain repaired point cloud data;
[0024] Extracting a characteristic curve from the repaired point cloud data;
[0025] Fitting the characteristic curve into a characteristic surface;
[0026] Based on the characteristic surface, a finite element model of the subdivided area is generated.
[0027] In a possible implementation manner of the first aspect, performing a force analysis on the finite element model to obtain a standard force value of the subdivided area includes:
[0028] Performing mesh division on the finite element model to obtain a divided mesh;
[0029] Identifying grid nodes in the partitioned grid;
[0030] Calculating nodal stresses of the grid nodes;
[0031] Identify the maximum value and the minimum value of the point cloud in the vertical direction of the divided grid;
[0032] Extracting a target point cloud on a point cloud path from the minimum point cloud value to the maximum point cloud value;
[0033] Based on the node stress, the point cloud stress of the target point cloud is calculated using the following formula:
[0034]
[0035] Among them, f j represents the point cloud stress, f i represents the node stress, o i represents the area contribution of the i-th grid node in the grid nodes to the target point cloud j, and n represents the total number of grid nodes in the grid nodes;
[0036] The node stress and the point cloud stress are used as the standard stress values.
[0037] In a possible implementation manner of the first aspect, calculating the node stress of the grid node includes:
[0038] The grid stress of the divided grid corresponding to the grid node is calculated using the following formula:
[0039]
[0040] Among them, (f 1 ,f 2 ,f 3 ,f 4 ) represents the node stress, f 1 Indicates the static pressure on the end face of the flange, f 2 It represents the force applied to the annular section between the sealing gasket and the inner diameter of the flange, f 3represents the force from external load, f 4 represents the compression force on the sealing gasket, a represents the force applied to the end face inside the flange in the horizontal direction, b represents the inner diameter of the flange, c represents the horizontal distance between the groove on the upper bottom surface inside the flange and the center of the flange, d represents the bending moment, f represents the preload force of the bolts in the holes of the flange, and e represents the axial traction force;
[0041] The mesh stress is taken as the nodal stress of the mesh nodes on the divided mesh.
[0042] In a possible implementation manner of the first aspect, determining the stress monitoring point of the flange based on the position-force influence relationship includes:
[0043] Extracting edge positions and standard force values in the position-force influence relationship;
[0044] When the standard force value at the edge position is similar to the standard force values at other edge positions among the edge positions, query the number of positions at other edge positions among the edge positions having standard force values similar to the standard force value at the edge position;
[0045] selecting a first target position from the edge positions using the number of positions;
[0046] Based on the position-force influence relationship, selecting a second target position affected by the local stress of the flange from the first target position;
[0047] Based on the position-force influence relationship, a third target position where no unit stress jitter of the flange occurs is selected from the second target position, and the third target position is used as the stress monitoring point.
[0048] In a possible implementation manner of the first aspect, calculating the intrinsic modal component of the acoustic emission signal includes:
[0049] The acoustic emission signal is modally decomposed using the following formula to obtain the initial modal components:
[0050]
[0051] Among them, p 1 represents the initial modal component, EMD represents the empirical mode decomposition algorithm, q represents the acoustic emission signal, λ represents white noise, x represents the number of times white noise is added to the acoustic emission signal, and X represents the total number of times white noise is added to the acoustic emission signal;
[0052] Based on the initial modal component, the residual of the acoustic emission signal is calculated using the following formula:
[0053] ε1 =qp 1
[0054] Among them, ε 1 represents the residual, p 1 represents the initial modal component, q represents the acoustic emission signal;
[0055] After the residual meets the preset conditions, calculating the fault parameters and related parameters of the initial modal component;
[0056] Calculating a mean square error of the initial modal component based on the fault parameter and the related parameters;
[0057] The intrinsic modal component is selected from the initial modal components using the mean square error.
[0058] In a possible implementation manner of the first aspect, extracting cepstrum parameters from the intrinsic modal components includes:
[0059] Performing component preprocessing on the connotation modal component to obtain a preprocessing component;
[0060] Identifying frequency domain components of the preprocessing components in the frequency domain;
[0061] constructing an energy spectrum of the frequency domain component;
[0062] The transfer function of energy spectrum filtering of the energy spectrum using a preset filter is calculated using the following formula:
[0063]
[0064] Among them, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, g(l-1) represents a preset filter, and l represents the Mel value obtained by performing energy spectrum filtering on the energy spectrum using the preset filter;
[0065] Based on the transfer function, the logarithmic energy of the filter is calculated using the following formula:
[0066]
[0067] Wherein, W(l) represents the logarithmic energy, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, l represents the Mel value obtained by filtering the energy spectrum using a preset filter, and D(k) represents the frequency domain component;
[0068] Performing cosine transformation on the logarithmic energy to obtain the cepstrum parameter.
[0069] In a possible implementation manner of the first aspect, the outputting the stress analysis value of the stress monitoring point by using the stress analysis model includes:
[0070] Normalizing the cepstrum parameters to obtain normalized parameters;
[0071] Based on the forget gate in the stress analysis model, the forget feature of the normalized parameter output by the forget gate in the stress analysis model is calculated using the following formula:
[0072] f t =σ(W f ·[h t-1 ,x t ]+b f )
[0073] Among them, f t represents the forgetting feature, σ represents the activation function, W f represents the weight of the forget gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t represents the normalization parameter, b f represents the bias of the forget gate, and t represents the time;
[0074] Calculating an input gate feature of the forgotten feature output by an input gate in the stress analysis model;
[0075] Based on the input gate feature, the final output feature in the stress analysis model is calculated using the following formula:
[0076] h t =o t *tanh(σ(W o [h t-1 ,x t ]+b o ))
[0077] Among them, h t represents the final output feature, o t represents the output result of the output gate in the stress analysis model, W o represents the weight of the output gate, b o represents the bias of the output gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t Denotes that the normalization parameter σ represents an activation function;
[0078] The stress analysis value of the stress monitoring point is identified using a classifier in the stress analysis model.
[0079] In a second aspect, the present invention provides a flange stress analysis system, the system comprising:
[0080] An image segmentation module, used for collecting a flange image of the flange, performing edge detection on the flange image to obtain a detected edge of the flange image, and performing image segmentation on the flange image using the detected edge to obtain a segmented image;
[0081] A cloud map construction module, used for identifying the subdivided area of the flange from the subdivided image, constructing a finite element model of the subdivided area, performing a force analysis on the finite element model to obtain a standard force value of the subdivided area, and constructing a force cloud map of the subdivided area based on the standard force value;
[0082] A monitoring point determination module, used to identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship;
[0083] A parameter extraction module, used for collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component;
[0084] The stress output module is used to input the cepstrum parameters into the stress analysis model, so as to output the stress analysis value of the stress monitoring point by using the stress analysis model, and use the stress analysis value as the stress analysis result of the flange.
[0085] Compared with the prior art, the technical principle and beneficial effects of this solution are:
[0086] The embodiment of the present invention performs edge detection on the flange image to identify the boundary of each object in the flange image, thereby determining the area where each object is located. Furthermore, the embodiment of the present invention performs image segmentation on the flange image by utilizing the detected edge, so as to divide the flange image into multiple images containing only one type of object based on different objects existing in the flange image. Furthermore, the embodiment of the present invention constructs a finite element model of the subdivided area to convert the solid flange into a three-dimensional virtual flange, so as to facilitate force value analysis of the flange from the three-dimensional virtual model. Furthermore, the embodiment of the present invention performs force analysis on the finite element model to analyze the idealized force condition of the flange using a finite element analysis method. By identifying the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, the stress change on the path from the starting point to the end point on each detection edge is analyzed. Furthermore, the embodiment of the present invention calculates the intrinsic modal component of the acoustic emission signal to eliminate the problem signal and noise in the acoustic emission signal and retain the key signal. Furthermore, the embodiment of the present invention extracts the cepstrum parameters from the intrinsic modal component to extract the features of the acoustic signal. The embodiment of the present invention outputs the stress analysis value of the stress monitoring point by using the stress analysis model to use the neural network model to intelligently output the corresponding stress analysis result based on the collected on-site acoustic emission signal, thereby improving the stress analysis accuracy of the flange on a practical basis. Therefore, the stress analysis method and system of a flange proposed in the embodiment of the present invention can improve the stress analysis accuracy of the flange on a practical basis. BRIEF DESCRIPTION OF THE DRAWINGS
[0087] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0088] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0089] Figure 1 A schematic flow chart of a flange stress analysis method provided by one embodiment of the present invention;
[0090] Figure 2a In one embodiment of the present invention Figure 1 A schematic diagram of a flange providing a method for stress analysis of a flange;
[0091] Figure 2b In one embodiment of the present invention Figure 1 A schematic diagram of node stress of a flange stress analysis method is provided;
[0092] Figure 3 In one embodiment of the present invention Figure 1 A schematic diagram of the position-force influence relationship of a flange stress analysis method is provided;
[0093] Figure 4 A module schematic diagram of a flange stress analysis system provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0094] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0095] The embodiment of the present invention provides a method for stress analysis of a flange, and the execution subject of the stress analysis method of the flange includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present invention. In other words, the stress analysis method of the flange can be executed by software or hardware installed in a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.
[0096] See also Figure 1 FIG. 1 is a flow chart of a flange stress analysis method provided by an embodiment of the present invention. Figure 1 The stress analysis methods for flanges described in include:
[0097] S1. Capturing a flange image of a flange, performing edge detection on the flange image to obtain a detected edge of the flange image, and performing image segmentation on the flange image using the detected edge to obtain a segmented image.
[0098] In the embodiments of the present invention, the flange refers to a part that connects shafts to each other and is used for connecting pipe ends. It is also used as a flange on the inlet and outlet of equipment and is used for connecting two devices, such as a reducer flange, also called a flange flange or a flange, etc.; the flange image refers to an image of the flange applied to an actual scene. In other words, in the actual application scenario of the flange, the flange image includes not only the flange, but also other objects connected to the flange.
[0099] Furthermore, the embodiment of the present invention performs edge detection on the flange image to identify the boundary of each object in the flange image, thereby determining the region where each object is located.
[0100] In one embodiment of the present invention, performing edge detection on the flange image to obtain the detected edge of the flange image includes: calculating the edge value of the flange image using the following formula:
[0101]
[0102] Wherein, G represents the edge value, G x represents the gradient value of filtering the flange image I in the horizontal direction, G y represents the gradient value of filtering the flange image I in the vertical direction;
[0103] When the edge value is greater than a preset edge value, the pixel point corresponding to the edge value in the flange image is used as the edge pixel point; the pixel point similarity between each edge pixel point in the edge pixel points and the neighboring pixel points of the edge pixel point is calculated; based on the pixel point similarity, each edge pixel point in the edge pixel points is connected pixel by pixel to obtain the detected edge.
[0104] The preset edge value may be set according to actual scenarios.
[0105] Optionally, based on the pixel similarity, each edge pixel in the edge pixel points is connected pixel by pixel, and the process of detecting the edge is as follows: when the pixel similarity between the edge pixel point and the neighborhood pixel point of the edge pixel point is greater than a preset similarity, it means that the edge pixel point and the neighborhood pixel point of the edge pixel point are relatively similar and can be used as pixel points on the boundary line of the same object. Therefore, the edge pixel point at this time and the neighborhood pixel point of the edge pixel point are connected to obtain the connected boundary line of the object.
[0106] Furthermore, the embodiment of the present invention performs image segmentation on the flange image by utilizing the detected edge, so as to divide the flange image into a plurality of images containing only one type of object based on different objects existing in the flange image.
[0107] Optionally, the process of performing image segmentation on the flange image by utilizing the detected edge to obtain a subdivided image is a process of dividing the flange image along the detected edge.
[0108] S2. Identify the segmented area of the flange from the segmented image, construct a finite element model of the segmented area, perform force analysis on the finite element model to obtain a standard force value of the segmented area, and construct a force cloud map of the segmented area based on the standard force value.
[0109] In the embodiment of the present invention, the subdivided area refers to a portion of the object in the flange corresponding to the object in the subdivided image that contains only one type of object.
[0110] Furthermore, the embodiment of the present invention constructs a finite element model of the subdivided area to convert the solid flange into a three-dimensional virtual flange, thereby facilitating force value analysis of the flange from the three-dimensional virtual model.
[0111] The finite element model refers to a three-dimensional model consisting of a curved surface fitted by point cloud data.
[0112] In one embodiment of the present invention, constructing the finite element model of the subdivided area includes: collecting point cloud data of the subdivided area; mapping the point cloud coordinates of the point cloud data to a two-dimensional coordinate system using the following formula to obtain two-dimensional coordinates:
[0113]
[0114] Wherein, (u, v) represents the two-dimensional coordinates, and (x, y, z) represents the point cloud coordinates of the point cloud data;
[0115] Based on the two-dimensional coordinates, the point cloud data is repaired to obtain the repaired point cloud data; characteristic curves are extracted from the repaired point cloud data; the characteristic curves are fitted into characteristic surfaces; and based on the characteristic surfaces, a finite element model of the subdivided area is generated.
[0116] Optionally, the process of repairing the point cloud data based on the two-dimensional coordinates to obtain the repaired point cloud data is: the above-mentioned process of mapping the point cloud coordinates of the point cloud data to the two-dimensional coordinate system is that the color information of each point of the three-dimensional object is stored in a 2D diffuse reflection coefficient table. Therefore, the process of an embodiment of the present invention is to calculate the corresponding coordinates of each point of the three-dimensional object in the diffuse reflection coefficient table according to the mapping relationship, and query the diffuse reflection coefficient corresponding to the calculated coordinate position. Based on this diffuse reflection coefficient, the texture color of each point of the three-dimensional object can be determined, thereby supplementing the originally missing texture color of the three-dimensional point cloud.
[0117] Furthermore, the embodiment of the present invention performs a stress analysis on the finite element model to analyze the idealized stress condition of the flange using a finite element analysis method.
[0118] In one embodiment of the present invention, the force analysis of the finite element model to obtain the standard force value of the subdivided area includes: meshing the finite element model to obtain a divided grid; identifying the grid nodes in the divided grid; calculating the node stress of the grid node; identifying the point cloud maximum value and the point cloud minimum value of the divided grid in the vertical direction; extracting the target point cloud on the point cloud path from the point cloud minimum value to the point cloud maximum value; based on the node stress, calculating the point cloud stress of the target point cloud using the following formula:
[0119]
[0120] Among them, f j represents the point cloud stress, f i represents the node stress, o i represents the area contribution of the i-th grid node in the grid nodes to the target point cloud j, and n represents the total number of grid nodes in the grid nodes;
[0121] The node stress and the point cloud stress are used as the standard stress values.
[0122] The divided grid is triangular in shape; the target point cloud refers to the point cloud within the divided grid, excluding the three nodes corresponding to the three corners of the divided grid.
[0123] In another embodiment of the present invention, the calculating the node stress of the grid node includes: calculating the grid stress of the divided grid corresponding to the grid node using the following formula:
[0124]
[0125] Among them, (f 1 ,f 2 ,f 3 ,f 4 ) represents the node stress, f 1 Indicates the static pressure on the end face of the flange, f 2 It represents the force applied to the annular section between the sealing gasket and the inner diameter of the flange, f 3 represents the force from external load, f 4 represents the compression force on the sealing gasket, a represents the force applied to the end face inside the flange in the horizontal direction, b represents the inner diameter of the flange, c represents the horizontal distance between the groove on the upper bottom surface inside the flange and the center of the flange, d represents the bending moment, f represents the preload force of the bolts in the holes of the flange, and e represents the axial traction force;
[0126] The mesh stress is taken as the nodal stress of the mesh nodes on the divided mesh.
[0127] See also Figure 2a As shown, an embodiment of the present invention Figure 1 A schematic diagram of a flange providing a method for stress analysis of a flange.
[0128] See also Figure 2b As shown, an embodiment of the present invention Figure 1 A schematic diagram of the node stress of a flange stress analysis method is provided. Figure 2b In the figure, a represents the force applied to the end face inside the flange in the horizontal direction, b represents the inner diameter of the flange, and c represents the horizontal distance between the groove on the upper bottom surface inside the flange and the center of the flange.
[0129] In the embodiment of the present invention, the force cloud diagram refers to a three-dimensional model diagram in which the force direction and force intensity are added to the finite element model.
[0130] S3. Identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship.
[0131] The embodiment of the present invention identifies the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map to analyze the change of stress on the path from the starting point to the end point on each detection edge.
[0132] In one embodiment of the present invention, determining the stress monitoring point of the flange based on the position-force influence relationship includes: extracting the edge position and the standard force value in the position-force influence relationship; when the standard force value at the edge position is similar to the standard force value of other edge positions among the edge positions, querying the number of positions of other edge positions among the edge positions having standard force values similar to the standard force value of the edge position; selecting a first target position from the edge positions using the number of positions; selecting a second target position affected by the local stress of the flange from the first target position based on the position-force influence relationship; selecting a third target position from the second target position where the unit stress jitter of the flange does not occur based on the position-force influence relationship, and using the third target position as the stress monitoring point.
[0133] It should be noted that the standard force value of the position not affected by the local stress in the position-force influence relationship is relatively stable, without obvious fluctuation. Furthermore, the reason for selecting the second target position affected by the local stress of the flange is that the transformation of the local stress can be queried from the standard force value; the standard force value of the third target position where unit stress jitter occurs in the position-force influence relationship will show sudden changes.
[0134] Exemplarily, the process of querying the number of other edge positions in the edge position having a standard force value similar to that of the edge position is as follows: edge A is composed of three edge segments a, b, and c, edge B is composed of two edge segments d and e, and edge C is composed of two edge segments f and h. When the standard force value of a is similar to the standard force values of d and e, and the standard force value of b is similar to the standard force values of f and h, the number of positions similar to the standard force value of edge A are edges B and C, and the number of positions is 2. It should be noted that other edge positions can only be included in the number of positions when all edge segments of other edge positions are similar to the standard force values of the edge position. The process of querying the number of other edge positions in the edge position having a standard force value similar to that of the edge position is as follows: since the above-mentioned edges B and C are similar to the standard force values of edge A, the standard force value of edge A can be used to replace the standard force values of edges B and C.
[0135] See also Figure 3 As shown, an embodiment of the present invention Figure 1 A schematic diagram of the position-force influence relationship of a flange stress analysis method is provided. Figure 3 In the figure, the horizontal axis is the grid division number, and the vertical axis is the stress value.
[0136] S4, collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component.
[0137] In the embodiment of the present invention, the acoustic emission signal refers to an ultra-high frequency stress wave pulse signal released due to the distortion of the molecular lattice, the intensification of cracks and the plastic deformation of the material during metal processing.
[0138] Furthermore, the embodiment of the present invention calculates the intrinsic modal components of the acoustic emission signal to eliminate problematic signals and noise in the acoustic emission signal and retain key signals.
[0139] The intrinsic modal components refer to the signal components of each layer obtained after the original signal is decomposed by empirical mode.
[0140] In one embodiment of the present invention, the calculating of the intrinsic modal components of the acoustic emission signal includes: performing modal decomposition on the acoustic emission signal using the following formula to obtain initial modal components:
[0141]
[0142] Among them, p 1represents the initial modal component, EMD represents the empirical mode decomposition algorithm, q represents the acoustic emission signal, λ represents white noise, x represents the number of times white noise is added to the acoustic emission signal, and X represents the total number of times white noise is added to the acoustic emission signal;
[0143] Based on the initial modal component, the residual of the acoustic emission signal is calculated using the following formula:
[0144] ε 1 =qp 1
[0145] Among them, ε 1 represents the residual, p 1 represents the initial modal component, q represents the acoustic emission signal;
[0146] After the residual meets the preset conditions, the fault parameters and related parameters of the initial modal component are calculated; based on the fault parameters and the related parameters, the mean square error of the initial modal component is calculated; and the intrinsic modal component is selected from the initial modal component using the mean square error.
[0147] Among them, the residual refers to the last component obtained by modal decomposition of the acoustic emission signal; the preset condition refers to the condition that the residual satisfies monotonicity and cannot be further decomposed. It should be noted that if the residual does not meet the preset condition, the acoustic emission signal continues to be decomposed, that is, the above-mentioned process of modal decomposition of the acoustic emission signal is returned; the fault parameter is used to indicate that there are more fault signals in the signal, and the calculation of the fault parameter can be achieved by calculating the kurtosis value; the correlation parameter is used to indicate the degree of correlation between the decomposed signal component and the original signal, which can be achieved by calculating the mutual correlation coefficient.
[0148] Optionally, the process of calculating the mean square error of the initial modal component based on the fault parameter and the related parameters is: when the fault parameter and the related parameters are both greater than the preset fault parameter and the preset related parameters, randomly selecting a target initial modal component from the initial modal components greater than the preset fault parameter and the preset related parameters, calculating the mean square error of the target initial modal component, and when the mean square error is minimum, using the target initial modal component as the intrinsic modal component.
[0149] Furthermore, the embodiment of the present invention extracts cepstrum parameters from the intrinsic modal components to extract features from the acoustic signal.
[0150] The cepstral parameter refers to the Mel cepstral coefficient.
[0151] In one embodiment of the present invention, the extracting cepstrum parameters from the intrinsic modal components includes: performing component preprocessing on the intrinsic modal components to obtain preprocessed components; identifying frequency domain components of the preprocessed components in the frequency domain; constructing an energy spectrum of the frequency domain components; and calculating a transfer function for filtering the energy spectrum using a preset filter using the following formula:
[0152]
[0153] Among them, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, g(l-1) represents a preset filter, and l represents the Mel value obtained by performing energy spectrum filtering on the energy spectrum using the preset filter;
[0154] Based on the transfer function, the logarithmic energy of the filter is calculated using the following formula:
[0155]
[0156] Wherein, W(l) represents the logarithmic energy, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, l represents the Mel value obtained by filtering the energy spectrum using a preset filter, and D(k) represents the frequency domain component;
[0157] Performing cosine transformation on the logarithmic energy to obtain the cepstrum parameter.
[0158] The preset filter is a Mel filter.
[0159] S5. Input the cepstrum parameters into a stress analysis model, so as to use the stress analysis model to output a stress analysis value of the stress monitoring point, and use the stress analysis value as a stress analysis result of the flange.
[0160] The embodiment of the present invention uses the stress analysis model to output the stress analysis value of the stress monitoring point, so as to use the neural network model to intelligently output the corresponding stress analysis result based on the collected on-site acoustic emission signal, thereby improving the stress analysis accuracy of the flange on a practical basis.
[0161] Wherein, the stress analysis model is trained by an LSTM neural network model.
[0162] In one embodiment of the present invention, the step of outputting the stress analysis value of the stress monitoring point using the stress analysis model includes: normalizing the cepstrum parameters to obtain normalized parameters; and calculating the forgetting characteristics of the normalized parameters output by the forgetting gate in the stress analysis model using the following formula based on the forgetting gate in the stress analysis model:
[0163] f t =σ(W f ·[h t-1 ,x t ]+b f )
[0164] Among them, f t represents the forgetting feature, σ represents the activation function, W f represents the weight of the forget gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t represents the normalization parameter, b f represents the bias of the forget gate, and t represents the time;
[0165] Calculate the input gate feature of the forgotten feature output by the input gate in the stress analysis model; and calculate the final output feature in the stress analysis model based on the input gate feature using the following formula:
[0166] h t =o t *tanh(σ(W o [h t-1 ,x t ]+b o ))
[0167] Among them, h t represents the final output feature, o t represents the output result of the output gate in the stress analysis model, W o represents the weight of the output gate, b o represents the bias of the output gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t Denotes that the normalization parameter σ represents an activation function;
[0168] The stress analysis value of the stress monitoring point is identified using a classifier in the stress analysis model.
[0169] It can be seen that the embodiment of the present invention performs edge detection on the flange image to identify the boundary of each object in the flange image, thereby determining the area where each object is located. Furthermore, the embodiment of the present invention performs image segmentation on the flange image by utilizing the detected edge, so as to divide the flange image into multiple images containing only one type of object based on different objects existing in the flange image. Furthermore, the embodiment of the present invention constructs a finite element model of the subdivided area to convert the solid flange into a three-dimensional virtual flange, so as to facilitate the force value analysis of the flange from the three-dimensional virtual model. Furthermore, the embodiment of the present invention performs force analysis on the finite element model to analyze the idealized force condition of the flange using the finite element analysis method. The embodiment of the present invention For example, by identifying the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, the stress change on the path from the starting point to the end point on each detection edge is analyzed. Furthermore, the embodiment of the present invention calculates the intrinsic modal component of the acoustic emission signal to eliminate the problem signal and noise in the acoustic emission signal and retain the key signal. Furthermore, the embodiment of the present invention extracts the cepstrum parameters from the intrinsic modal component to extract the features of the acoustic signal. The embodiment of the present invention outputs the stress analysis value of the stress monitoring point by using the stress analysis model, so as to use the neural network model to intelligently output the corresponding stress analysis result based on the collected on-site acoustic emission signal, thereby improving the stress analysis accuracy of the flange on a practical basis. Therefore, a flange stress analysis method proposed in an embodiment of the present invention can improve the stress analysis accuracy of the flange on a practical basis.
[0170] like Figure 4 The figure shows a functional module diagram of the stress analysis system of the flange of the present invention.
[0171] The flange stress analysis system 400 of the present invention can be installed in an electronic device. According to the functions to be implemented, the flange stress analysis system can include an image segmentation module 401, a cloud map construction module 402, a monitoring point determination module 403, a parameter extraction module 404 and a stress output module 405. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.
[0172] In the embodiment of the present invention, the functions of each module / unit are as follows:
[0173] The image segmentation module 401 is used to collect a flange image of the flange, perform edge detection on the flange image to obtain a detected edge of the flange image, and perform image segmentation on the flange image using the detected edge to obtain a subdivided image;
[0174] The cloud map construction module 402 is used to identify the subdivided area of the flange from the subdivided image, construct a finite element model of the subdivided area, perform force analysis on the finite element model to obtain a standard force value of the subdivided area, and construct a force cloud map of the subdivided area based on the standard force value;
[0175] The monitoring point determination module 403 is used to identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship;
[0176] The parameter extraction module 404 is used to collect the acoustic emission signal at the stress monitoring point, calculate the intrinsic modal component of the acoustic emission signal, and extract the cepstrum parameter from the intrinsic modal component;
[0177] The stress output module 405 is used to input the cepstrum parameters into the stress analysis model, so as to output the stress analysis value of the stress monitoring point using the stress analysis model, and use the stress analysis value as the stress analysis result of the flange.
[0178] In detail, the modules in the flange stress analysis system 400 in the embodiment of the present invention are used in the same manner as described above. Figures 1 to 3 The stress analysis method of the flange described in the invention is the same technical means and can produce the same technical effect, so it will not be repeated here.
[0179] The present invention further provides a storage medium, wherein the readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:
[0180] Acquire a flange image of the flange, perform edge detection on the flange image to obtain a detected edge of the flange image, and perform image segmentation on the flange image using the detected edge to obtain a segmented image;
[0181] Identify the segmented area of the flange from the segmented image, construct a finite element model of the segmented area, perform force analysis on the finite element model to obtain a standard force value of the segmented area, and construct a force cloud map of the segmented area based on the standard force value;
[0182] Identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship;
[0183] Collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component;
[0184] The cepstrum parameters are input into a stress analysis model, so as to use the stress analysis model to output stress analysis values of the stress monitoring points, and the stress analysis values are used as stress analysis results of the flange.
[0185] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.
[0186] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0187] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0188] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0189] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.
[0190] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0191] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
Claims
1. A flange stress analysis method, characterized in that: The method comprises: Acquire a flange image of the flange, perform edge detection on the flange image to obtain a detected edge of the flange image, and perform image segmentation on the flange image using the detected edge to obtain a segmented image; Identify the segmented area of the flange from the segmented image, construct a finite element model of the segmented area, perform force analysis on the finite element model to obtain a standard force value of the segmented area, and construct a force cloud map of the segmented area based on the standard force value; Identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship; Collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component; The cepstrum parameters are input into a stress analysis model, so as to use the stress analysis model to output stress analysis values of the stress monitoring points, and the stress analysis values are used as stress analysis results of the flange.
2. The method according to claim 1, characterized in that The performing edge detection on the flange image to obtain the detected edge of the flange image includes: The edge value of the flange image is calculated using the following formula: Wherein, G represents the edge value, G x represents the gradient value of filtering the flange image I in the horizontal direction, G y represents the gradient value of filtering the flange image I in the vertical direction; When the edge value is greater than a preset edge value, a pixel point corresponding to the edge value in the flange image is used as an edge pixel point; Calculating the pixel similarity between each edge pixel in the edge pixels and the neighboring pixels of the edge pixel; Based on the pixel similarity, each edge pixel in the edge pixel is connected pixel by pixel to obtain the detected edge.
3. The method according to claim 1, characterized in that The step of constructing the finite element model of the subdivided area comprises: Collecting point cloud data of the subdivided area; The point cloud coordinates of the point cloud data are mapped to a two-dimensional coordinate system using the following formula to obtain two-dimensional coordinates: Wherein, (u, v) represents the two-dimensional coordinates, and (x, y, z) represents the point cloud coordinates of the point cloud data; Based on the two-dimensional coordinates, repair the point cloud data to obtain repaired point cloud data; Extracting a characteristic curve from the repaired point cloud data; Fitting the characteristic curve into a characteristic surface; Based on the characteristic surface, a finite element model of the subdivided area is generated.
4. The method according to claim 1, characterized in that The performing stress analysis on the finite element model to obtain the standard stress value of the subdivided area includes: Performing mesh division on the finite element model to obtain a divided mesh; Identifying grid nodes in the partitioned grid; Calculating nodal stresses of the grid nodes; Identify the maximum value and the minimum value of the point cloud in the vertical direction of the divided grid; Extracting a target point cloud on a point cloud path from the minimum point cloud value to the maximum point cloud value; Based on the node stress, the point cloud stress of the target point cloud is calculated using the following formula: Among them, f j represents the point cloud stress, f i represents the node stress, o i represents the area contribution of the i-th grid node in the grid nodes to the target point cloud j, and n represents the total number of grid nodes in the grid nodes; The node stress and the point cloud stress are used as the standard stress values.
5. The method according to claim 4, characterized in that The calculating the node stress of the grid node comprises: The grid stress of the divided grid corresponding to the grid node is calculated using the following formula: Wherein, (f1, f2, f3, f4) represents the node stress, f1 represents the static pressure on the end face inside the flange, f2 represents the force applied to the annular section between the sealing gasket and the inner diameter of the flange, f3 represents the force from the external load, f4 represents the compressive force from the sealing gasket, a represents the force applied to the end face inside the flange in the horizontal direction, b represents the inner diameter of the flange, c represents the horizontal distance between the groove on the upper bottom surface inside the flange and the center of the flange, d represents the bending moment, f represents the preload force of the bolts in the hole of the flange, and e represents the axial traction force; The mesh stress is taken as the nodal stress of the mesh nodes on the divided mesh.
6. The method according to claim 1, characterized in that The step of determining the stress monitoring point of the flange based on the position-force influence relationship includes: Extracting edge positions and standard force values in the position-force influence relationship; When the standard force value at the edge position is similar to the standard force values at other edge positions among the edge positions, query the number of positions at other edge positions among the edge positions having standard force values similar to the standard force value at the edge position; selecting a first target position from the edge positions using the number of positions; Based on the position-force influence relationship, selecting a second target position affected by the local stress of the flange from the first target position; Based on the position-force influence relationship, a third target position where no unit stress jitter of the flange occurs is selected from the second target position, and the third target position is used as the stress monitoring point.
7. The method according to claim 1, characterized in that The calculating of the intrinsic modal component of the acoustic emission signal comprises: The acoustic emission signal is modally decomposed using the following formula to obtain the initial modal components: Wherein, p1 represents the initial modal component, EMD represents the empirical mode decomposition algorithm, q represents the acoustic emission signal, λ represents white noise, x represents the number of times white noise is added to the acoustic emission signal, and X represents the total number of times white noise is added to the acoustic emission signal; Based on the initial modal component, the residual of the acoustic emission signal is calculated using the following formula: ε1=q-p1 Wherein, ε1 represents the residual, p1 represents the initial modal component, and q represents the acoustic emission signal; After the residual meets the preset conditions, calculating the fault parameters and related parameters of the initial modal component; Calculating a mean square error of the initial modal component based on the fault parameter and the related parameters; The intrinsic modal component is selected from the initial modal components using the mean square error.
8. The method according to claim 1, characterized in that The step of extracting cepstrum parameters from the connotation modal components comprises: Performing component preprocessing on the connotation modal component to obtain a preprocessing component; Identifying frequency domain components of the preprocessing components in the frequency domain; constructing an energy spectrum of the frequency domain component; The transfer function of energy spectrum filtering of the energy spectrum using a preset filter is calculated using the following formula: Among them, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, g(l-1) represents a preset filter, and l represents the Mel value obtained by performing energy spectrum filtering on the energy spectrum using the preset filter; Based on the transfer function, the logarithmic energy of the filter is calculated using the following formula: Wherein, W(l) represents the logarithmic energy, H l (k) represents the transfer function, k represents the frequency index in the energy spectrum, l represents the Mel value obtained by filtering the energy spectrum using a preset filter, and D(k) represents the frequency domain component; Performing cosine transformation on the logarithmic energy to obtain the cepstrum parameter.
9. The method according to claim 1, characterized in that: The step of outputting the stress analysis value of the stress monitoring point by using the stress analysis model comprises: Normalizing the cepstrum parameters to obtain normalized parameters; Based on the forget gate in the stress analysis model, the forget feature of the normalized parameter output by the forget gate in the stress analysis model is calculated using the following formula: f t =σ(W f ·[h t-1 ,x t ]+b f ) Among them, f t represents the forgetting feature, σ represents the activation function, W f represents the weight of the forget gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t represents the normalization parameter, b f represents the bias of the forget gate, and t represents the time; Calculating an input gate feature of the forgotten feature output by an input gate in the stress analysis model; Based on the input gate feature, the final output feature in the stress analysis model is calculated using the following formula: h t =o t *tanh(σ(W o [h t-1 ,x t ]+b o )) Among them, h t represents the final output feature, o t represents the output result of the output gate in the stress analysis model, W o represents the weight of the output gate, b o represents the bias of the output gate, h t-1 Represents the output data of the stress analysis model at the last moment, x t Denotes that the normalization parameter σ represents an activation function; The stress analysis value of the stress monitoring point is identified using a classifier in the stress analysis model.
10. A flange stress analysis system, characterized in that: The system comprises: An image segmentation module, used for collecting a flange image of the flange, performing edge detection on the flange image to obtain a detected edge of the flange image, and performing image segmentation on the flange image using the detected edge to obtain a segmented image; A cloud map construction module, used for identifying the subdivided area of the flange from the subdivided image, constructing a finite element model of the subdivided area, performing a force analysis on the finite element model to obtain a standard force value of the subdivided area, and constructing a force cloud map of the subdivided area based on the standard force value; A monitoring point determination module, used to identify the position-force influence relationship between the detection edge corresponding to the subdivided area and the standard force value in the force cloud map, and determine the stress monitoring point of the flange based on the position-force influence relationship; A parameter extraction module, used for collecting the acoustic emission signal at the stress monitoring point, calculating the intrinsic modal component of the acoustic emission signal, and extracting cepstrum parameters from the intrinsic modal component; The stress output module is used to input the cepstrum parameters into the stress analysis model, so as to output the stress analysis value of the stress monitoring point by using the stress analysis model, and use the stress analysis value as the stress analysis result of the flange.
Citation Information
Patent Citations
A wind turbine tower stress state calculation method based on a modal superposition method
CN109543258A
Automatic modeling method of high-fidelity finite element model for bladed disc structure
CN114970247A
Power equipment inspection method and device
CN116168351A
Electrical equipment noise source point identification method and system based on imaging technology
CN117171627A
Cited By
Method for evaluating stress of column foot node of concrete filled steel tubular column for waste incineration power generation
CN119885402A
