Welding seam type identification method, device and equipment and storage medium

Abstract

The invention relates to the technical field of weld joint recognition, and discloses a weld joint type recognition method, device and equipment and a storage medium. The method comprises the following steps: acquiring a to-be-identified grid model; classifying the to-be-identified welding seam according to the target welding seam feature to obtain a first classification label; determining the type of the to-be-identified welding seam according to the first classification label; when the type of the to-be-recognized weld joint is not determined according to the first classification label, inputting the target model feature and the target weld joint feature into a pre-trained weld joint type recognition model, and classifying the to-be-recognized weld joint according to the target model feature and the target weld joint feature to obtain a second classification label; and determining the type of the to-be-identified welding seam according to the second classification label. According to the embodiment of the invention, the welding seam type identification efficiency and precision can be improved.

Description

Technical Field

[0001] This application relates to the technical field of weld identification, and in particular to a method, device, equipment and storage medium for identifying weld types. Background Art

[0002] Welds are high-incidence areas of stress concentration and fatigue failure in engineering structures. The assessment of their fatigue life is directly related to the safety and economy of equipment. Whether the fatigue analysis of weld structures can be carried out quickly and efficiently is of great significance.

[0003] In actual engineering scenarios, such as the fields of automobiles and construction machinery, there are usually some complex structural parts. These complex structural parts mix T-shaped welds and lap welds. The prior art requires manual classification of weld types during modeling, and setting material parameters and fatigue parameters for each part. However, the requirements for defining weld types for each part and modeling weld elements strictly make the process of weld fatigue analysis trivial and complicated. Manual classification and repeated parameter setting are time-consuming, and cannot meet the needs of rapid iterative design, affecting the efficiency and accuracy of weld structure fatigue life assessment and engineering decision-making. Summary of the Invention

[0004] The purpose of this application is to provide a method, device, equipment and storage medium for identifying weld types, aiming to improve the efficiency and accuracy of weld type identification.

[0005] An embodiment of this application provides a method for identifying weld types, including: Obtain a mesh model to be identified; the mesh model to be identified contains welds to be identified; Classify the welds to be identified according to target weld features to obtain a first classification label; the target weld features include the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plates where the welds to be identified are located; Determine the type of the welds to be identified according to the first classification label; When the type of the welds to be identified cannot be determined according to the first classification label, input the target model features and the target weld features into a pre-trained weld type identification model, and classify the welds to be identified according to the target model features and the target weld features to obtain a second classification label; the target model features include the model topology feature and the model application feature of the mesh model to be identified; Determine the type of the welds to be identified according to the second classification label.

[0006] In one embodiment, before classifying the welds to be identified according to the target weld features, it further includes: Obtain the coordinate information of both the welding surface and the plate; Calculate the normal vector of the welding surface based on the coordinate information of the welding surface, and determine the angular feature of the welding surface based on the normal vector of the welding surface; Determine the connection topology of the welding surface based on the coordinate information of the plate, and determine the connection feature of the welding surface based on the connection topology.

[0007] In one embodiment, the classifying the weld to be recognized according to the target weld feature includes: Input the target weld feature into a static classifier, and classify the weld to be recognized according to the angle range where the angle corresponding to the angular feature of the welding surface is located and the connection type of the connection topology of the welding surface corresponding to the connection feature of the welding surface, to obtain the first classification label.

[0008] In one embodiment, the classifying the weld to be recognized according to the target model feature and the target weld feature includes: Splice the target model feature and the target weld feature to obtain a spliced feature; the model topology feature in the target model feature includes the structural assembly feature, welding surface topology feature, welding surface material feature, and welding surface metallographic feature of the weld to be recognized grid model; Map the spliced feature to a feature space to obtain a mapped feature; Classify the mapped feature to obtain a number of initial prediction labels and corresponding confidence levels; Determine the initial prediction label corresponding to the confidence level with the largest value as the second classification label.

[0009] In one embodiment, the classifying the mapped feature includes: Determine a weight value according to the matching degree between the target model feature and the target weld feature; Use the classification tree of the weld type recognition model to classify the mapped feature to generate the initial prediction label; Determine the confidence level corresponding to the initial prediction label according to the weight value.

[0010] In one embodiment, the weld type recognition method further includes: When receiving a weld type recognition operation, input the target model feature and the target weld feature into the weld type recognition model, and classify the weld to be recognized according to the target model feature and the target weld feature to obtain the second classification label; Determine the type of the weld to be recognized according to the first classification label or the second classification label.

[0011] In one embodiment, the weld type recognition method further includes: Evaluate the model topology performance of the to-be-recognized mesh model according to the target weld seam features and the target model features, and generate corresponding prompt information according to the model topology performance evaluation result.

[0012] The embodiment of the present application also provides a weld seam type recognition device, including: The first module is used to obtain a to-be-recognized mesh model; the to-be-recognized mesh model contains to-be-recognized weld seams; The second module is used to classify the to-be-recognized weld seams according to the target weld seam features to obtain a first classification label; the target weld seam features include the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plate members where the to-be-recognized weld seam is located; The third module is used to determine the type of the to-be-recognized weld seam according to the first classification label; The fourth module is used to, when the type of the to-be-recognized weld seam cannot be determined according to the first classification label, input the target model features and the target weld seam features into a pre-trained weld seam type recognition model, and classify the to-be-recognized weld seam according to the target model features and the target weld seam features to obtain a second classification label; the target model features include the model topology feature and the model application feature of the to-be-recognized mesh model; The fifth module is used to determine the type of the to-be-recognized weld seam according to the second classification label.

[0013] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the above-mentioned weld seam type recognition method is implemented.

[0014] The embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned weld seam type recognition method is implemented.

[0015] The beneficial effects of the present application: It is not necessary to manually classify the to-be-recognized weld seams. Instead, the to-be-recognized weld seams are classified according to the target weld seam features of the to-be-recognized weld seams to obtain a first classification label, so as to determine the type of the to-be-recognized weld seam according to the first classification label. When the type of the to-be-recognized weld seam cannot be determined according to the first classification label, the target model features of the to-be-recognized mesh model and the target weld seam features of the to-be-recognized weld seam are input into a pre-trained weld seam type recognition model, and the to-be-recognized weld seam is classified according to the target model features and the target weld seam features to obtain a second classification label, so as to determine the type of the to-be-recognized weld seam according to the second classification label. By adaptively and automatically identifying the type of the to-be-recognized weld seam, the efficiency and accuracy of weld seam type recognition can be improved. Description of the Drawings

[0016] Figure 1 It is a flowchart of the weld type recognition method provided by an embodiment of the present application.

[0017] Figure 2 It is a flowchart of the method before step S102 provided by an embodiment of the present application.

[0018] Figure 3 It is a flowchart of the specific method of step S104 provided by an embodiment of the present application.

[0019] Figure 4 It is a flowchart of the specific method of step S303 provided by an embodiment of the present application.

[0020] Figure 5 It is a schematic structural diagram of the weld type recognition device provided by an embodiment of the present application.

[0021] Figure 6 It is a schematic hardware structure diagram of the electronic device provided by an embodiment of the present application. Detailed implementation manners

[0022] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0023] It should be noted that although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown can be executed in a different order from the module division in the device or the order in the flowchart. Terms such as "first" and "second" in the description, claims and drawings are used to distinguish similar objects and not to describe a specific order or sequence.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0025] Figure 1 It is a flowchart of a weld type recognition method provided by an embodiment of the present application. Refer to Figure 1 , in one embodiment, the method includes but is not limited to steps S101 to S105.

[0026] Step S101, obtain the grid model to be recognized.

[0027] The grid model to be recognized contains the weld to be recognized.

[0028] The grid model to be recognized is stored in a grid model file, which refers to the best model file saved after hyperparameter tuning through grid search, or a file containing multiple hyperparameter combinations and their corresponding model performances. Grid search is a hyperparameter optimization method that trains multiple models and evaluates their performances by traversing all predefined hyperparameter combinations (i.e., the "grid"), and finally selects the optimal hyperparameter combination. The grid model to be recognized is a grid model generated for industrial products and contains several groups of welds to be recognized, such as components of an automotive chassis or aerospace appliances, etc.

[0029] Specifically, to obtain the grid model to be recognized, it can be to read the grid model file generated from CAE software, and then parse the coordinate information, type information, connection relationship information, and attribute information of each plate, welding surface, and node in the grid model to be recognized, and store the parsed information as multiple information sets.

[0030] Step S102, classify the welds to be recognized according to the target weld features to obtain the first classification label.

[0031] The target weld features include the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plates where the weld to be recognized is located. The target weld features can be obtained by extracting the information related to the weld to be recognized from the information sets obtained by parsing various types of information in the grid model to be recognized, including the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plates where the weld to be recognized is located. Among them, the welding surface angle feature refers to the angle feature between the welding surfaces of the two plates where the weld to be recognized is located, and the welding surface connection feature refers to the connection feature of the connection relationship between the welding surfaces of the two plates where the weld to be recognized is located.

[0032] Specifically, classifying the welds to be recognized according to the target weld features can be to compare the target weld features with each weld feature in the preset weld feature rule library to determine the feature space where the target weld features are located, so as to determine the type of the weld to be recognized according to the comparison result and assign the corresponding classification label. The weld feature rule library can be configured as a static classifier. Input the target weld features into the static classifier, and the static classifier outputs the first classification label representing the type of the weld to be recognized.

[0033] Figure 2 It is a flowchart of the method before step S102 provided by the embodiments of the present application. Refer to Figure 2 In one embodiment, the method includes but is not limited to steps S201 to S203.

[0034] Step S201, obtain the coordinate information of both the welding surface and the plate.

[0035] Specifically, the coordinate information of both the welding surface and the plate can be extracted from the information set obtained by analyzing various types of information in the mesh model to be recognized.

[0036] Step S202: Calculate the normal vector of the welding surface based on the coordinate information of the welding surface, and determine the angle feature of the welding surface based on the normal vector of the welding surface.

[0037] Specifically, to determine the angle feature of the welding surface, a mapping relationship between each welding surface and other welding surfaces can be established through the connection relationship between the welding surfaces. Then, the normal vector of the welding surface is calculated based on the coordinate information of the welding surface, and the angle feature of the welding surface is determined based on the mapping relationship between the welding surface and other welding surfaces and the normal vector of the welding surface. In a specific embodiment, the calculation formulas for both the normal vector of the welding surface and the angle feature of the welding surface are as follows: , , , where, is the angle feature of the welding surface, and are the normal vectors of the welding surface, , , and are the unit edge vectors of the welding surface.

[0038] Step S203: Determine the connection topology of the welding surface based on the coordinate information of the plate, and determine the connection feature of the welding surface based on the connection topology.

[0039] Specifically, to determine the connection feature of the welding surface, a mapping relationship between each plate and other plates can be established through the connection relationship between the plates. Then, the connection topology of the welding surface is determined based on the coordinate information of the plate and the mapping relationship between the plate and other plates, and the angle feature of the welding surface is determined based on the mapping relationship between the plate and other plates and the connection topology of the welding surface.

[0040] In one embodiment, classifying the welds to be recognized according to the target weld features includes: inputting the target weld features into a static classifier, and classifying the welds to be recognized according to the angle interval in which the angle corresponding to the angle feature of the welding surface is located and the connection type of the connection topology of the welding surface corresponding to the connection feature of the welding surface to obtain a first classification label.

[0041] Specifically, the weld to be recognized is classified by using a weld feature rule library. The angle range where the angle corresponding to the welding surface angle feature is located is matched with the corresponding angle range in the weld feature rule library, and the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature is matched with the corresponding connection type in the weld feature rule library. The weld to be recognized is classified by combining the two matching results, so as to obtain the corresponding first classification label. For example, when the angle corresponding to the welding surface angle feature is in the range of 45° to 90°, if the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature is a spatial orthogonal distribution (that is, the included angle is close to 90° and the central planes of the two plates intersect), then this weld is determined to be a T-shaped weld, and a first classification label indicating that the weld to be recognized is a T-shaped weld is obtained. When the angle corresponding to the welding surface angle feature is in the range less than 45°, if the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature is a plate overlap joint (by calculating that the ratio of the projected overlapping area of the two plates to the area of the smaller plate is greater than a preset threshold, usually 30%), then this weld is determined to be an overlap joint weld, and a first classification label indicating that the weld to be recognized is an overlap joint weld is obtained. When the angle corresponding to the welding surface angle feature is close to 180°, if the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature is that the center lines of the two plates are basically coplanar, then this weld is determined to be a butt weld, and a first classification label indicating that the weld to be recognized is a butt weld is obtained. When the angle corresponding to the welding surface angle feature is in the range of 60° to 120°, if the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature is that the two plates share a common edge, then this weld is determined to be a fillet weld, and a first classification label indicating that the weld to be recognized is a fillet weld is obtained.

[0042] Step S103, determine the type of the weld to be recognized according to the first classification label.

[0043] Specifically, the first classification label is parsed to extract the type information of the weld to be recognized in the first classification label, so as to obtain the type of the weld to be recognized. If the static classifier fails to match the angle range where the angle corresponding to the welding surface angle feature is located and / or the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature in the weld feature rule library, then the obtained first classification label does not contain the type information of the weld to be recognized, and the type of the weld to be recognized cannot be determined according to the first classification label.

[0044] Step S104, when the type of the weld to be recognized cannot be determined according to the first classification label, input the target model feature and the target weld feature into a pre-trained weld type recognition model, and classify the weld to be recognized according to the target model feature and the target weld feature to obtain a second classification label.

[0045] The target model features include the model topology features and model application features of the mesh model to be recognized. The target model features can be obtained by extracting information related to the topological structure of the mesh model to be recognized from the information set obtained by parsing various types of information in the mesh model to be recognized, including model topology features and model application features. Among them, the model topology features refer to the features of the topological structure of the mesh model to be recognized, and the model application features refer to the features of the application scenario of the mesh model to be recognized.

[0046] Specifically, classifying the weld to be recognized according to the target model features and target weld features can be to use a pre-trained weld type recognition model to classify the target model features and target weld features in multiple dimensions based on the learned engineering cases, so as to determine the type of the weld to be recognized carrying the target weld features under the corresponding target model feature conditions, so as to obtain the corresponding second classification label. In the embodiment of the present application, the weld type recognition model uses the random forest algorithm to classify the weld to be recognized, and the weld type recognition model is trained based on the training feature set composed of the sample model features and sample weld features and the control feature set composed of the control model features and control weld features corresponding to the engineering cases.

[0047] Figure 3 It is a flowchart of the specific method of step S104 provided in the embodiment of the present application. Refer to Figure 3 In one embodiment, the method includes but is not limited to steps S301 to S304.

[0048] Step S301, splice the target model features and target weld features to obtain the spliced features.

[0049] In this embodiment, the model topology features in the target model features include the structural assembly features, welding surface topology features, welding surface material features, and welding surface metallographic features of the mesh model to be recognized.

[0050] Step S302, map the spliced features to the feature space to obtain the mapped features.

[0051] Taking the spliced features as the input data of the feature mapping layer of the weld type recognition model, in the process of processing by the feature mapping layer, first map the spliced features to the feature space to obtain the mapped features. The mapped features are usually in vector form for subsequent processing.

[0052] Step S303, classify the mapped features to obtain a number of initial prediction labels and corresponding confidence levels.

[0053] Here, the mapping features can be optimized to use more effective information as the mapping features. The optimization process can be convolution processing or integration processing, etc. The way of optimization is determined according to the type of the weld type recognition model. Then, the optimized mapping features are classified to obtain at least one initial prediction label and the confidence corresponding to each initial prediction label. Among them, the confidence represents the credibility of the initial prediction label.

[0054] Step S304: Determine the initial prediction label corresponding to the confidence with the largest value as the second classification label.

[0055] Determine the initial prediction label corresponding to the confidence with the largest value as the prediction label of the image classification model. For example, when classifying the mapping features, an initial prediction label is 0 with a corresponding confidence of 0.3, and another initial prediction label is 1 with a corresponding confidence of 0.7. Therefore, the initial prediction label with a value of 1 is determined as the second classification label.

[0056] Figure 4 It is a flowchart of the specific method of step S303 provided by the embodiment of the present application. Refer to Figure 4 In one embodiment, the method includes but is not limited to steps S401 to S403.

[0057] Step S401: Determine the weight value according to the matching degree between the target model feature and the target weld feature.

[0058] When determining the matching degree between the target model feature and the target weld feature, it can be to calculate the distances between the target model feature and the target weld feature and the cluster center point, and determine the matching degree between the target model feature and the target weld feature according to the corresponding relationship between the distance and the matching degree. The higher the matching degree, the larger the weight value, and the lower the matching degree, the smaller the weight value.

[0059] Step S402: Use the classification tree of the weld type recognition model to classify the mapping features and generate initial prediction labels.

[0060] The mapping features are classified using the classification tree of the weld type recognition model. First, the mapping features can be clustered to obtain the clusters to which the mapping features belong. After determining the clusters to which the mapping features belong, the classification tree corresponding to the clusters of the mapping features can be selected according to the correspondence between the clusters and the classification tree for classification calculation, and the classification to which the mapping features belong can be obtained through the classification tree calculation. Since the classification trees in the random forest in this application are constructed based on the clustered sample subsets, when classifying the welds to be recognized, the correspondence between the clusters and the classification trees can be determined according to the clustering of the sample subsets when constructing the classification trees. Classifying and calculating through the classification tree corresponding to the cluster to which the mapping features belong can effectively improve the classification calculation efficiency, and since the mapping features are similar to the sample subsets of the constructed classification trees, the accuracy of the classification can be effectively guaranteed.

[0061] Step S403: Determine the confidence level corresponding to the initial prediction label according to the weight value.

[0062] Determining the confidence level corresponding to the initial prediction label according to the weight value can be done after obtaining multiple initial prediction labels based on the classification tree of the weld type recognition model, and fusing the weight values with the same initial prediction label, for example, by summing, to obtain the confidence level of the initial prediction label.

[0063] Step S105: Determine the type of the weld to be recognized according to the second classification label.

[0064] Specifically, the second classification label is parsed to extract the type information of the weld to be recognized in the second classification label, and the type of the weld to be recognized is obtained.

[0065] In one embodiment, the weld type recognition method further includes: when receiving a weld type recognition operation, inputting the target model features and the target weld features into the weld type recognition model, classifying the weld to be recognized according to the target model features and the target weld features to obtain a second classification label; determining the type of the weld to be recognized according to the first classification label or the second classification label.

[0066] Specifically, after determining the type of the weld to be recognized according to the first classification label, the user can obtain the second classification label by performing a weld type recognition operation. When the execution entity receives the weld type recognition operation, it inputs the target model features and the target weld features into the weld type recognition model, classifies the weld to be recognized according to the target model features and the target weld features, and obtains the second classification label. After obtaining the second classification label, the type of the weld to be recognized can be determined according to the first classification label or the second classification label.

[0067] In one embodiment, the weld type recognition method further includes: evaluating the model topological performance of the mesh model to be recognized based on the target weld features and the target model features, and generating corresponding prompt information according to the model topological performance evaluation result.

[0068] Specifically, first, after traversing all weld surfaces and their adjacent weld surfaces around, evaluate the mesh quality of these weld surfaces, mark and temporarily store the weld surfaces whose mesh quality is lower than the critical requirements. Next, based on the obtained target weld features and target model features, determine whether each weld surface meets the weld modeling requirements. If not, also record these weld surfaces, summarize these potentially problematic weld surfaces, and generate corresponding prompt information. The generated prompt information includes the specific problems existing in these weld surfaces and corresponding optimization methods.

[0069] Please refer to Figure 5 , the embodiment of the present application also provides a weld type recognition device, which can implement the above weld type recognition method. The device includes: The first module 501 is used to obtain the mesh model to be recognized; the mesh model to be recognized contains the welds to be recognized; The second module 502 is used to classify the welds to be recognized according to the target weld features to obtain the first classification label; the target weld features include the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plate members where the welds to be recognized are located; The third module 503 is used to determine the type of the welds to be recognized according to the first classification label; When the type of the welds to be recognized cannot be determined according to the first classification label, the fourth module 504 inputs the target model features and the target weld features into the pre-trained weld type recognition model, classifies the welds to be recognized according to the target model features and the target weld features, and obtains the second classification label; the target model features include the model topological features and the model application features of the mesh model to be recognized; The fifth module 505 is used to determine the type of the welds to be recognized according to the second classification label.

[0070] The specific implementation manner of this weld type recognition device is basically the same as the specific embodiments of the above weld type recognition method, and will not be elaborated here.

[0071] Figure 6 is a block diagram of an electronic device shown according to an exemplary embodiment.

[0072] Next, refer to Figure 6 to describe the electronic device 600 according to this embodiment of the present disclosure. Figure 6 The electronic device 600 shown is only an example, and should not bring any limitation to the functions and usage scope of the embodiments of the present disclosure.

[0073] As shown Figure 6 in the figure, the electronic device 600 is presented in the form of a general computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different system components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0074] Among them, the storage unit stores program code, and the program code can be executed by the processing unit 610, so that the processing unit 610 executes the steps according to various exemplary embodiments of the present disclosure described in the above-mentioned weld type recognition method part of this specification.

[0075] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 6201 and / or a cache storage unit 6202, and may further include a read-only storage unit (ROM) 6203.

[0076] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205. Such program modules 6205 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0077] The bus 630 may represent one or more of several types of bus structures, including a storage unit bus or a storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any bus structure in a variety of bus structures.

[0078] The electronic device 600 can also communicate with one or more external devices 600' (such as a keyboard, a pointing device, a Bluetooth device, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication can be carried out through an input / output (I / O) interface 650. And, the electronic device 600 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 660. The network adapter 660 can communicate with other modules of the electronic device 600 through the bus 630. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0079] An embodiment of the present application also provides a computer-readable storage medium storing a computer program, which when executed by a processor implements the above-mentioned weld type recognition method.

[0080] The weld type recognition method, device, equipment and storage medium provided by the embodiments of the present application do not require manual classification of the welds to be recognized. Instead, the welds to be recognized are classified based on the target weld features of the welds to be recognized to obtain a first classification label, and the type of the welds to be recognized is determined based on the first classification label. When the type of the welds to be recognized cannot be determined based on the first classification label, the target model features of the mesh model to be recognized and the target weld features of the welds to be recognized are input into a pre-trained weld type recognition model, and the welds to be recognized are classified based on the target model features and the target weld features to obtain a second classification label, and the type of the welds to be recognized is determined based on the second classification label. By adaptively and automatically recognizing the type of the welds to be recognized, the efficiency and accuracy of weld type recognition can be improved.

[0081] Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions for causing a computing device (which can be a personal computer, a server, or a network device, etc.) to execute the above methods according to the embodiments of the present disclosure.

[0082] The program product can adopt any combination of one or more readable media. The readable media can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0083] A computer-readable storage medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The readable storage medium may also be any readable medium other than the readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0084] Those skilled in the art can understand that the above-mentioned modules can be distributed in the device according to the description of the embodiments, or can be correspondingly changed and be located in one or more devices that are uniquely different from this embodiment. The modules of the above embodiments can be combined into one module, or can be further split into multiple sub-modules.

[0085] The exemplary embodiments of the present disclosure have been specifically shown and described above. It should be understood that the present disclosure is not limited to the detailed structures, settings, or implementation methods described herein; on the contrary, the present disclosure is intended to cover various modifications and equivalent settings included within the spirit and scope of the appended claims.

Claims

1. A method for identifying weld types, characterized in that, Including: Obtain the grid model to be recognized; The grid model to be recognized includes welds to be recognized; Classify the welds to be recognized according to the target weld features to obtain the first classification label; The target weld features include the welding surface angle feature and the welding surface connection feature between the welding surfaces of the two plate parts where the welds to be recognized are located; Determine the type of the welds to be recognized according to the first classification label; When the type of the welds to be recognized cannot be determined according to the first classification label, input the target model features and the target weld features into a pre-trained weld type recognition model, and classify the welds to be recognized according to the target model features and the target weld features to obtain the second classification label; the target model features include the model topology feature and the model application feature of the grid model to be recognized; Determine the type of the welds to be recognized according to the second classification label.

2. The weld type recognition method according to claim 1, characterized in that Before classifying the welds to be recognized according to the target weld features, it further includes: Obtain the coordinate information of both the welding surface and the plate part; Calculate the normal vector of the welding surface according to the coordinate information of the welding surface, and determine the welding surface angle feature according to the normal vector of the welding surface; Determine the connection topology of the welding surface according to the coordinate information of the plate part, and determine the welding surface connection feature according to the connection topology.

3. The weld type recognition method according to claim 1, characterized in that, The classifying the welds to be recognized according to the target weld features includes: Input the target weld features into a static classifier, and classify the welds to be recognized according to the angle interval where the angle corresponding to the welding surface angle feature is located and the connection type of the connection topology of the welding surface corresponding to the welding surface connection feature to obtain the first classification label.

4. The weld type recognition method according to claim 1, characterized in that The classifying the welds to be recognized according to the target model features and the target weld features includes: Splice the target model features and the target weld features to obtain the spliced features; the model topology feature in the target model features includes the structural assembly feature, the welding surface topology feature, the welding surface material feature, and the welding surface metallographic feature of the grid model to be recognized; Map the spliced features to the feature space to obtain the mapped features; Classify the mapped features to obtain a number of initial prediction labels and the corresponding confidence levels; Determine the initial prediction label corresponding to the confidence level with the largest value as the second classification label.

5. The weld type recognition method according to claim 4, wherein The classifying the mapped features includes: Determine the weight value according to the matching degree between the target model features and the target weld features; Use the classification tree of the weld type recognition model to classify the mapped features to generate the initial prediction labels; Determine the confidence level corresponding to the initial prediction label according to the weight value.

6. The weld type recognition method according to claim 1, wherein It further includes: When receiving a weld type recognition operation, input the target model features and the target weld features into the weld type recognition model, and classify the welds to be recognized according to the target model features and the target weld features to obtain the second classification label; Determine the type of the welds to be recognized according to the first classification label or the second classification label.

7. The weld type recognition method according to claim 1, characterized in that Further included are: evaluating the model topology performance of the to-be-identified mesh model according to the target weld feature and the target model feature, and generating corresponding prompt information according to the model topology performance evaluation result.

8. A weld type recognition device, characterized in that, Including: a first module for obtaining a to-be-identified mesh model; the to-be-identified mesh model contains a to-be-identified weld; a second module for classifying the to-be-identified weld according to the target weld feature to obtain a first classification label; the target weld feature includes a weld surface angle feature and a weld surface connection feature between the welding surfaces of two plate members where the to-be-identified weld is located; a third module for determining the type of the to-be-identified weld according to the first classification label; a fourth module for, when the type of the to-be-identified weld cannot be determined according to the first classification label, inputting the target model feature and the target weld feature into a pre-trained weld type recognition model, and classifying the to-be-identified weld according to the target model feature and the target weld feature to obtain a second classification label; the target model feature includes a model topology feature and a model application feature of the to-be-identified mesh model; a fifth module for determining the type of the to-be-identified weld according to the second classification label.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the weld type recognition method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the weld type recognition method according to any one of claims 1 to 7 is implemented.

Images