Method, device and equipment for identifying fracture toughness and brittleness of metal and medium

Abstract

The invention provides a metal fracture ductile-brittleness identification method and device, equipment and a medium, relates to the technical field of material failure analysis, and aims to solve the problem of how to accurately identify metal fracture ductile-brittleness. The metal fracture toughness and brittleness identification method comprises the steps of obtaining a to-be-predicted microscopic feature image data set of a metal fracture surface; inputting the to-be-predicted microscopic feature image data set into a pre-trained microscopic feature recognition model to determine the number corresponding to each microscopic feature contained in the to-be-predicted microscopic feature image data set; and inputting the number corresponding to each microscopic feature contained in the to-be-predicted microscopic feature image data set into a pre-trained regression prediction model to determine the ductile brittleness of the metal fracture.

Description

Technical Field

[0001] This application relates to the field of materials failure analysis technology, and in particular to a method, apparatus, equipment and medium for identifying the ductile-brittle fracture properties of metals. Background Technology

[0002] Identifying the fracture surface type of metals plays a crucial role in materials failure analysis, serving as a primary task and a key support for industrial systems and materials science. Determining the fracture surface type of a metal can optimize materials development pathways, avoid excessive engineering costs, provide a basis for industry standard setting, and balance safety and efficiency.

[0003] However, the fracture surface type of metals needs to be determined by experts based on experience and other factors, and even experts may not be able to reach a consensus. Therefore, accurately identifying the ductile-brittle fracture characteristics of metals is a problem that urgently needs to be solved. Summary of the Invention

[0004] This disclosure provides a method, apparatus, device, and medium for identifying the brittle fracture properties of metals, aiming to solve the problem of how to accurately identify the brittle fracture properties of metals.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] In a first aspect, a method for identifying the ductile-brittle fracture properties of metals is provided, comprising: acquiring a dataset of microscopic features to be predicted on the fracture surface of the metal; inputting the dataset of microscopic features to be predicted into a pre-trained microscopic feature recognition model to determine the quantity of each microscopic feature contained in the dataset of microscopic features to be predicted; and inputting the quantity of each microscopic feature contained in the dataset of microscopic features to be predicted into a pre-trained regression prediction model to determine the ductile-brittle fracture properties of the metal.

[0007] In some embodiments, the pre-trained micro-feature recognition model is determined by: obtaining a first training sample set; the first training sample set includes multiple first training samples and a label for each first training sample; the first training samples include a first micro-feature image dataset; the labels of the first training samples include micro-features corresponding to the first micro-feature image dataset; the first micro-feature image dataset is an image set containing a single type of micro-feature; based on the first training sample set, the original deep residual network model is trained to obtain a single-type micro-feature classification model; obtaining a second training sample set, the second training sample set includes multiple second training samples and a label for each second training sample; the second training samples include a second micro-feature image dataset; the labels of the second training samples include multiple types of micro-features corresponding to the second micro-feature image dataset; the second micro-feature image dataset is an image set containing multiple types of micro-features; based on the second training sample set, the original deep residual network model is trained and combined with the single-type micro-feature classification model to obtain the pre-trained micro-feature recognition model.

[0008] In some embodiments, the label of each second training sample is determined by: inputting the second micro-feature image dataset into a single-class micro-feature classification model, determining multiple single-class micro-features contained in the second micro-feature image dataset, and determining the multi-class micro-features corresponding to the second micro-feature image dataset; the multi-class micro-features include multiple single-class micro-features; and determining the multi-class micro-features contained in the second micro-feature image dataset as the label of the second training sample.

[0009] In some embodiments, the pre-trained regression prediction model is determined by: obtaining a training sample set; the training sample set includes: multiple training samples, prediction features of each training sample, and a target to be predicted for each training sample; the training samples include a first micro-feature image dataset and a second micro-feature image dataset; the prediction features of the training samples include the quantity corresponding to each micro-feature contained in the first micro-feature image dataset and the quantity corresponding to each micro-feature contained in the second micro-feature image dataset; the target to be predicted for the training samples includes ductile fracture and brittle fracture; based on the training sample set, the original extreme gradient boosting tree model is trained to obtain the pre-trained regression prediction model.

[0010] In some embodiments, the pre-trained regression prediction model satisfies the following formula:

[0011]

[0012] Where x is the dataset of microscopic feature images to be predicted; η represents the ductility and brittleness of metal fracture corresponding to the dataset of microscopic feature images to be predicted; T represents the total number of decision trees in the model; η represents the learning rate; q represents the ductility and brittleness of metal fracture. t (x) is the index of the leaf node to which the micro-feature image dataset to be predicted is mapped in the t-th tree; The value corresponding to the index of the leaf node.

[0013] In some embodiments, the microstructure includes at least one of dimples, intergranular structures, and cleavage.

[0014] In some embodiments, acquiring a dataset of images of the microscopic features to be predicted on the fracture surface of a metal includes: acquiring the dataset of images of the microscopic features to be predicted on the fracture surface of a metal by scanning electron microscopy.

[0015] In a second aspect, a device for identifying the brittle fracture properties of metals is provided, the device comprising: a communication unit and a processing unit;

[0016] The communication unit is used to acquire a dataset of images of the microscopic features to be predicted on the fracture surface of the metal.

[0017] The processing unit is used to input the micro-feature image dataset to be predicted into a pre-trained micro-feature recognition model to determine the quantity of each micro-feature contained in the micro-feature image dataset to be predicted.

[0018] The processing unit is also used to input the quantity corresponding to each micro-feature contained in the micro-feature image dataset to be predicted into a pre-trained regression prediction model to determine the ductility and brittleness of metal fracture.

[0019] Thirdly, a device for identifying the brittle fracture properties of metals is provided, including a memory and a processor; the memory is used to store computer-executed instructions, and the processor is connected to the memory via a bus; when the device for identifying the brittle fracture properties of metals is running, the processor executes the computer-executed instructions stored in the memory, so that the device for identifying the brittle fracture properties of metals performs the method for identifying the brittle fracture properties of metals in the first aspect.

[0020] The device for identifying the brittle fracture properties of metals can be an electronic device or a component of an electronic device, such as a chip system within an electronic device. The chip system supports the electronic device in implementing the functions involved in the first aspect and any possible implementation thereof, such as acquiring and determining the data and / or information involved in the aforementioned method for identifying the brittle fracture properties of metals. The chip system includes a chip, but may also include other discrete devices or circuit structures.

[0021] Fourthly, a computer-readable storage medium is provided, comprising computer-executable instructions that, when executed on a computer, cause the computer to perform the method for identifying the brittle fracture characteristics of metals described in the first aspect.

[0022] Fifthly, a computer program product is also provided, comprising a computer program or instructions that, when executed on a metal fracture brittleness identification device, cause the metal fracture brittleness identification device to perform the metal fracture brittleness identification method as described in the first aspect above.

[0023] It should be noted that the aforementioned computer instructions may be stored, in whole or in part, on a computer-readable storage medium. This computer-readable storage medium may be packaged together with the processor of the metal fracture brittleness identification device, or it may be packaged separately from the processor of the metal fracture brittleness identification device; this application does not limit this.

[0024] The descriptions of the second, third, fourth, and fifth aspects of this application can be referenced to the detailed description of the first aspect.

[0025] In the embodiments of this application, the name of the aforementioned metal fracture brittleness identification device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear under other names. For example, the receiving unit may also be called a receiving module, receiver, etc. As long as the function of each device or functional module is similar to that of this application, it falls within the scope of the claims of this application and its equivalents.

[0026] This application provides a method for identifying the ductile-brittle fracture properties of metals. It involves acquiring a dataset of images of microscopic features to be predicted on the fracture surface of the metal. Next, the dataset of images of the microscopic features to be predicted can be input into a pre-trained microscopic feature recognition model to determine the quantity of each microscopic feature contained in the dataset. Then, the quantity of each microscopic feature contained in the dataset can be input into a pre-trained regression prediction model to determine the ductile-brittle fracture properties of the metal.

[0027] As shown above, this scheme accurately classifies micro-features through a pre-trained micro-feature recognition model, achieving the identification of the type and quantity of micro-features. Then, the classification results (i.e., the quantity of each micro-feature in the dataset of images to be predicted) are input into a pre-trained regression prediction model. Through the analysis of the type, quantity, and proportion of micro-features by the regression prediction model, and based on the inherent correlation between the ductility and brittleness of metal fracture and micro-features (i.e., ductile fracture corresponds to a large number of dimples, and brittle fracture corresponds to cleavage or intergranular fracture), the ductility and brittleness of metal fracture are accurately inferred, fully considering the inherent characteristics of the material and exploring the performance of the metal fracture surface itself. This scheme, by combining the micro-feature recognition model and the regression prediction model, makes the identification of the ductility and brittleness of metal fracture less susceptible to disturbances from the external environment. Therefore, this scheme comprehensively solves the problem of how to accurately identify the ductility and brittleness of metal fracture. Attached Figure Description

[0028] Figure 1 A schematic diagram of the structure of a metal fracture ductility and brittleness identification system provided in this application embodiment;

[0029] Figure 2 A schematic diagram of the hardware structure of a metal fracture brittleness identification device provided in this application embodiment;

[0030] Figure 3 A flowchart illustrating a method for identifying the ductile-brittle fracture properties of metals provided in this application embodiment;

[0031] Figure 4 A flowchart illustrating a method for determining a fracture toughness / brittleness identification model, provided as an embodiment of this application;

[0032] Figure 5 A function graph of the performance index of a regression prediction model provided in an embodiment of this application;

[0033] Figure 6 This is a schematic diagram of the structure of a metal fracture brittleness identification device provided in an embodiment of this application. Detailed Implementation

[0034] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0035] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0036] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0037] As described in the background section, identifying the fracture surface type of metals plays a crucial role in materials failure analysis. It is the primary task of materials failure analysis and a key support for industrial systems and materials science. Determining the fracture surface type of metals can optimize materials research and development pathways, avoid excessive engineering costs, provide a basis for industry standard setting, and balance safety and efficiency.

[0038] However, the fracture type of a metal needs to be determined by experts based on their experience and other factors, and even experts may not be able to reach a consensus.

[0039] Currently, a deep learning-based method for identifying metal fracture types has been disclosed, using images of metal fracture surfaces as input. The deep learning network employed constructs a Gaussian Process (GP) model and an Upper Confidence Bound (UCB) function, while introducing an attention mechanism into the network structure to complete model creation. However, this method neglects in-depth analysis of the performance of the metal fracture surface itself.

[0040] A three-dimensional graph learning-based method for metal fracture identification is also disclosed. This method involves acquiring a three-dimensional image of the metal fracture cross-section, smoothing the image to obtain a smooth surface, and then performing operations such as cropping, projection, and flattening on the smooth surface to obtain cross-sections that can serve as nodes. The corresponding connected region cross-sectional features and the fractal dimension features of each connected region are extracted from these cross-sections to construct a node feature matrix and a neighbor matrix. A graph convolutional neural network is then used to process these matrices to identify the metal fracture type. However, this method involves overly complex processing of the original image, and the processing measures only consider the geometric morphology of the fracture, failing to adequately consider the material's inherent properties. Furthermore, the selected feature parameters are easily affected by external environmental disturbances.

[0041] To address the aforementioned problems, this application provides a method for identifying the ductile-brittle fracture properties of metals. This method acquires a dataset of microscopic features to be predicted on the fracture surface of the metal. Next, the dataset is input into a pre-trained microscopic feature recognition model to determine the quantity of each microscopic feature contained in the dataset. Then, the quantity of each microscopic feature in the dataset is input into a pre-trained regression prediction model to determine the ductile-brittle fracture properties of the metal.

[0042] As shown above, this scheme accurately classifies micro-features through a pre-trained micro-feature recognition model, achieving the identification of the type and quantity of micro-features. Then, the classification results (i.e., the quantity of each micro-feature in the dataset of images to be predicted) are input into a pre-trained regression prediction model. Through the analysis of the type, quantity, and proportion of micro-features by the regression prediction model, and based on the inherent correlation between the ductility and brittleness of metal fracture and micro-features (i.e., ductile fracture corresponds to a large number of dimples, and brittle fracture corresponds to cleavage or intergranular fracture), the ductility and brittleness of metal fracture are accurately inferred, fully considering the inherent characteristics of the material and exploring the performance of the metal fracture surface itself. This scheme, by combining the micro-feature recognition model and the regression prediction model, makes the identification of the ductility and brittleness of metal fracture less susceptible to disturbances from the external environment. Therefore, this scheme comprehensively solves the problem of how to accurately identify the ductility and brittleness of metal fracture.

[0043] The implementation environment of the above-mentioned method for identifying the brittle fracture properties of metals can be the metal fracture brittle fracture identification system provided in the embodiments of this application.

[0044] Figure 1 This is a schematic diagram of a metal fracture ductility and brittleness identification system provided in an embodiment of this application. Figure 1 As shown, the metal fracture ductility identification system includes: a metal fracture ductility identification device 101 and a data storage device 102.

[0045] The metal fracture ductility identification device 101 and the data storage device 102 are connected in communication.

[0046] In practical applications, the metal fracture ductility and brittleness identification device 101 can be connected to any number of data storage devices 102. For ease of understanding, Figure 1 The following is an example of a metal fracture brittleness identification device 101 connected to a data storage device 102.

[0047] In this embodiment of the application, the data storage device 102 is used to provide data for the identification of metal fracture brittleness (e.g., a dataset of microscopic feature images to be predicted) to the metal fracture brittleness identification device 101, so that the metal fracture brittleness identification device 101 can identify the metal fracture brittleness based on the data sent by the data storage device 102.

[0048] Optionally, the physical devices of the metal fracture ductility identification device 101 and the data storage device 102 can be servers, terminals, or other types of electronic devices, and this application embodiment does not limit them.

[0049] Optionally, the aforementioned terminal may be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing device connected to a wireless modem. The wireless terminal may communicate with one or more core networks via a radio access network (RAN). The wireless terminal may be a mobile terminal, such as a mobile phone (or "cellular" phone) and a computer with a mobile terminal, or a portable, pocket-sized, handheld, computer-embedded, or vehicle-mounted mobile device that exchanges voice and / or data with the radio access network, such as a mobile phone, tablet computer, laptop computer, netbook, or personal digital assistant (PDA).

[0050] Optionally, the server mentioned above can be one of the servers in a server cluster (composed of multiple servers), a chip in the server, a system-on-a-chip in the server, or a virtual machine (VM) deployed on a physical machine. This application embodiment does not limit this.

[0051] Optionally, the metal fracture ductility and brittleness identification device 101 and the data storage device 102 can be two independently configured devices, or they can be integrated into the same device. When the metal fracture ductility and brittleness identification device 101 and the data storage device 102 are integrated into the same device, the data storage device 102 can be a storage module (e.g., a database) of the metal fracture ductility and brittleness identification device 101.

[0052] It is easy to understand that when the metal fracture brittleness identification device 101 and the data storage device 102 are integrated into the same device, the communication method between the metal fracture brittleness identification device 101 and the data storage device 102 is the same as the communication between internal modules of the device. In this case, the communication process between the two is the same as when the metal fracture brittleness identification device 101 and the data storage device 102 are independent of each other.

[0053] For ease of understanding, this application will use the example of a metal fracture ductility and brittleness identification device 101 and a data storage device 102 operating independently.

[0054] The identification devices for metal fracture ductility and brittleness in a metal fracture identification system include, for example: Figure 2 The components included. The following are examples. Figure 2 Taking the metal fracture brittleness identification device shown as an example, the hardware structure of the metal fracture brittleness identification device is introduced.

[0055] Figure 2 This is a schematic diagram of the hardware structure of a metal fracture brittleness identification device provided in an embodiment of this application. Figure 2 As shown, the metal fracture ductility and brittleness identification device includes: a processor 201, a memory 202, a communication interface 203, and a bus 204. The processor 201, the memory 202, and the communication interface 203 can be connected via the bus 204.

[0056] Processor 201 is the control center of the metal fracture ductility and brittleness identification device. It can be a single processor or a collective term for multiple processing elements. For example, processor 201 can be a general-purpose central processing unit (CPU) or other general-purpose processors. Among them, the general-purpose processor can be a microprocessor or any conventional processor.

[0057] As one embodiment, processor 201 may include one or more CPUs, for example Figure 2 CPU0 and CPU1 are shown in the diagram.

[0058] The memory 202 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0059] In one possible implementation, the memory 202 can exist independently of the processor 201. The memory 202 can be connected to the processor 201 via a bus 204 and is used to store instructions or program code. When the processor 201 calls and executes the instructions or program code stored in the memory 202, it can implement the metal fracture ductility and brittleness identification method provided in the following embodiments of this application.

[0060] In this embodiment, the software programs stored in memory 202 differ for the metal fracture brittleness identification device, resulting in different functions implemented by the device. The functions performed by each device will be described in conjunction with the following flowchart.

[0061] In another possible implementation, the memory 202 can also be integrated with the processor 201.

[0062] The communication interface 203 is used for connecting the metal fracture ductility and brittleness identification device to other devices via a communication network, such as Ethernet, wireless access network, or wireless local area network (WLAN). The communication interface 203 may include a receiving unit for receiving data and a transmitting unit for transmitting data.

[0063] Bus 204 can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 2 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0064] It should be pointed out that, Figure 2 The structure shown does not constitute a limitation on the identification device for the brittle fracture properties of metals, except Figure 2 In addition to the components shown, the metal fracture brittleness identification device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0065] The method for identifying the brittle fracture properties of metals provided in this application will be described in detail below with reference to the accompanying drawings.

[0066] The method for identifying the brittle fracture properties of metals provided in this application is applied to... Figure 1The metal fracture ductility identification system shown includes a metal fracture ductility identification device 101, such as... Figure 3 As shown in the embodiments of this application, a method for identifying the brittle fracture ductility of metals includes:

[0067] S301, The metal fracture brittleness identification device acquires a dataset of images of microscopic features to be predicted on the fracture surface of the metal.

[0068] Specifically, in order to identify the type of microscopic features on the fracture surface of a metal, it is necessary to first obtain a dataset of images of the microscopic features to be predicted on the fracture surface of the metal.

[0069] Specifically, the types and proportions of microscopic features of different fracture surfaces directly correspond to whether the metal material undergoes a large amount of plastic deformation and consumes a large amount of energy before fracture. Therefore, the toughness and brittleness of metal fracture can be inferred from the types and proportions of microscopic feature images on the fracture surface of the metal.

[0070] For example, in some embodiments, the microstructure includes at least one of dimples, intergranular structures, and cleavage.

[0071] Optionally, microscopic features may also include fatigue.

[0072] For example, when the amount of plastic deformation is large and the energy consumption is high, the corresponding microstructure is dimples, and the fracture is ductile fracture. When the amount of plastic deformation is small and the energy consumption is low, the corresponding microstructure is cleavage, and the fracture is brittle fracture. When the amount of plastic deformation is even smaller and the energy consumption is low, the corresponding microstructure may be intergranular, and the fracture may be brittle fracture.

[0073] S302, the metal fracture brittleness identification device inputs the micro-feature image dataset to be predicted into a pre-trained micro-feature recognition model to determine the quantity of each micro-feature contained in the micro-feature image dataset to be predicted.

[0074] Specifically, in order to determine the ductility and brittleness of metal fracture, it is necessary to determine the number of each micro-feature contained in the image dataset of the micro-features to be predicted.

[0075] For example, in some embodiments, the pre-trained micro-feature recognition model is determined by: obtaining a first training sample set; the first training sample set includes multiple first training samples and a label for each first training sample; the first training samples include a first micro-feature image dataset; the labels of the first training samples include micro-features corresponding to the first micro-feature image dataset; the first micro-feature image dataset is an image set containing a single type of micro-feature; based on the first training sample set, the original deep residual network model is trained to obtain a single-type micro-feature classification model; obtaining a second training sample set, the second training sample set includes multiple second training samples and a label for each second training sample; the second training samples include a second micro-feature image dataset; the labels of the second training samples include multiple types of micro-features corresponding to the second micro-feature image dataset; the second micro-feature image dataset is an image set containing multiple types of micro-features; based on the second training sample set, the original deep residual network model is trained and combined with the single-type micro-feature classification model to obtain the pre-trained micro-feature recognition model.

[0076] For example, the first training sample set includes: a dimple dataset, a grain-side dataset, and a cleavage dataset. The second training sample set consists of multiple types of microscopic feature datasets.

[0077] Optionally, during training, the single-class micro-feature classification model and the micro-feature recognition model can be adjusted in real time based on the loss function to obtain the best-performing single-class micro-feature classification model and micro-feature recognition model.

[0078] As can be seen from the above embodiments, by training the original deep residual network model with a dataset containing a single type of micro-feature, a model that can identify a single type of micro-feature is trained. Then, by training the original deep residual network model with a dataset containing multiple types of micro-features and combining it with a single type of micro-feature classification model, a model that can identify multiple types of micro-features is further obtained, making the identification of micro-feature types more accurate.

[0079] S303, the metal fracture brittleness identification device inputs the quantity of each micro-feature contained in the micro-feature image dataset to be predicted into a pre-trained regression prediction model to determine the brittleness of metal fracture.

[0080] Specifically, in order to determine the ductility and brittleness of metal fracture, the output of the micro-feature recognition model needs to be input into the regression prediction model to determine the ductility and brittleness of metal fracture.

[0081] Specifically, the ductile-brittle fracture of metals includes both ductile fracture and brittle fracture.

[0082] For example, in some embodiments, the pre-trained regression prediction model is determined by: obtaining a training sample set; the training sample set includes: multiple training samples, prediction features of each training sample, and a target to be predicted for each training sample; the training samples include a first micro-feature image dataset and a second micro-feature image dataset; the prediction features of the training samples include the quantity corresponding to each micro-feature contained in the first micro-feature image dataset and the quantity corresponding to each micro-feature contained in the second micro-feature image dataset; the target to be predicted for the training samples includes ductile fracture and brittle fracture; based on the training sample set, the original extreme gradient boosting tree model is trained to obtain the pre-trained regression prediction model.

[0083] Specifically, the first microscopic feature image and the second microscopic feature image together constitute the dataset of the regression prediction model.

[0084] For example, in the structured data of the regression prediction model, ductile fracture is treated as a value of 1 and brittle fracture is treated as a value of 0.

[0085] For example, in some embodiments, the pre-trained regression prediction model satisfies the following formula:

[0086]

[0087] Where x is the dataset of microscopic feature images to be predicted; η represents the ductility and brittleness of metal fracture corresponding to the dataset of microscopic feature images to be predicted; T represents the total number of decision trees in the model; η represents the learning rate; q represents the ductility and brittleness of metal fracture. t (x) is the index of the leaf node to which the micro-feature image dataset to be predicted is mapped in the t-th tree; The value corresponding to the index of the leaf node.

[0088] As demonstrated in the above embodiments, the number of various micro-features contained in the microscopic feature image of the metal fracture surface is used to analyze and determine whether the metal fractures ductilely or brittlely. This method deeply mines information from the metal fracture surface itself, fully considers the material's inherent characteristics, and is not easily affected by external interference.

[0089] Figure 4 This is a schematic flowchart illustrating a method for determining a fracture toughness / brittleness identification model, provided as an embodiment of this application. Figure 4 As shown, the model for identifying fracture ductility and brittleness includes:

[0090] S401. Obtain microscopic feature images.

[0091] Specifically, the device for identifying the brittleness and toughness of metal fractures acquires a first microscopic feature image and a second microscopic feature image for training the model.

[0092] S402. Create a dataset.

[0093] Specifically, the metal fracture brittleness identification device annotates the first microscopic feature image and divides it into a training dataset, a validation dataset, and a test dataset in an 8:1:1 ratio to obtain the first training sample set.

[0094] S403, Training a single-category micro-feature classification model.

[0095] Specifically, the metal fracture brittleness identification device trains the original deep residual network model based on the first training sample set to obtain a single-category micro-feature classification model.

[0096] S404, Training the micro-feature recognition model.

[0097] Specifically, the metal fracture brittleness identification device annotates the second microscopic feature image using a single-category microscopic feature classification model, and divides it into a training dataset, a validation dataset, and a test dataset in an 8:1:1 ratio to determine the second training sample set. Then, the original deep residual network model is trained based on the second training sample set to obtain the microscopic feature recognition model.

[0098] S405, Training the regression prediction model.

[0099] Specifically, the metal fracture ductility and brittleness identification device trains the original extreme gradient boosting tree model based on the first microscopic feature image and the second microscopic feature image to obtain a regression prediction model.

[0100] S406. Determine the fracture ductility-brittleness identification model.

[0101] Specifically, the metal fracture ductility identification device uses a micro-feature identification model and a training regression prediction model as the fracture ductility identification model.

[0102] The Receiver Operating Characteristic Curve (ROC) can be used as a performance metric for the model, and the best-performing completed regression prediction model can be saved.

[0103] Figure 5 This is a function graph of the performance index of a regression prediction model provided in an embodiment of this application. For example... Figure 5As shown, the horizontal axis represents the false positive rate for negative samples, ranging from 0 to 1; the vertical axis represents the recognition rate for positive samples, also ranging from 0 to 1. The area under the ROC curve (AUC) reflects the model's discriminative ability. The AUC value of the regression prediction model is 0.9724, indicating that the model has strong discriminative ability and is effective in distinguishing the ductility and brittleness of metal fractures.

[0104] In some embodiments, the label of each second training sample is determined by: inputting the second micro-feature image dataset into a single-class micro-feature classification model, determining multiple single-class micro-features contained in the second micro-feature image dataset, and determining the multi-class micro-features corresponding to the second micro-feature image dataset; the multi-class micro-features include multiple single-class micro-features; and determining the multi-class micro-features contained in the second micro-feature image dataset as the label of the second training sample.

[0105] Specifically, in order to more accurately label the second training sample set, it is necessary to identify the second micro-feature image through a single-class micro-feature classification model to obtain the second training sample set.

[0106] In some embodiments, in S301 above, acquiring the image dataset of the microscopic features to be predicted on the fracture surface of the metal specifically includes:

[0107] The device for identifying the brittleness and toughness of metal fractures acquires a dataset of images of the microscopic features to be predicted on the fracture surface of the metal using a scanning electron microscope.

[0108] Specifically, in order to obtain images of the microscopic features to be predicted on the fracture surface of a metal, the fracture surface of the metal can be scanned using a scanning electron microscope.

[0109] Optionally, pipe metal materials obtained under different conditions can be acquired to ensure the diversity of raw materials and improve the generalization performance of the model.

[0110] For example, during the acquisition process, it is ensured that the images originate from the source region, the extended region, and the transient region, while also ensuring the randomness of the acquisition points.

[0111] Optionally, before scanning the metal, the acquired industrial metal material needs to be pretreated under laboratory conditions to avoid the influence of contaminants on the subsequent acquisition of microscopic feature images. Furthermore, an appropriate method should be selected for sample cleaning of the pipeline metal material based on the actual contamination status to avoid over-cleaning and causing damage to the sample surface.

[0112] The foregoing mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, it includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0113] This application embodiment can divide the metal fracture ductility and brittleness identification device into functional modules according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. Optionally, the module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0114] Figure 6 A schematic diagram of the structure of a metal fracture brittleness identification device provided in an embodiment of this application is shown. Figure 6 As shown, the metal fracture ductility identification device includes: a communication unit 601 and a processing unit 602;

[0115] The communication unit 601 is used to acquire a dataset of images of the microscopic features to be predicted on the fracture surface of the metal.

[0116] The processing unit 602 is used to input the micro-feature image dataset to be predicted into a pre-trained micro-feature recognition model to determine the quantity of each micro-feature contained in the micro-feature image dataset to be predicted.

[0117] The processing unit 602 is also used to input the quantity corresponding to each micro feature contained in the micro feature image dataset to be predicted into a pre-trained regression prediction model to determine the ductility and brittleness of metal fracture.

[0118] In some embodiments, the pre-trained micro-feature recognition model is determined by: obtaining a first training sample set; the first training sample set includes multiple first training samples and a label for each first training sample; the first training samples include a first micro-feature image dataset; the labels of the first training samples include micro-features corresponding to the first micro-feature image dataset; the first micro-feature image dataset is an image set containing a single type of micro-feature; based on the first training sample set, the original deep residual network model is trained to obtain a single-type micro-feature classification model; obtaining a second training sample set, the second training sample set includes multiple second training samples and a label for each second training sample; the second training samples include a second micro-feature image dataset; the labels of the second training samples include multiple types of micro-features corresponding to the second micro-feature image dataset; the second micro-feature image dataset is an image set containing multiple types of micro-features; based on the second training sample set, the original deep residual network model is trained and combined with the single-type micro-feature classification model to obtain the pre-trained micro-feature recognition model.

[0119] In some embodiments, the label of each second training sample is determined by: inputting the second micro-feature image dataset into a single-class micro-feature classification model, determining multiple single-class micro-features contained in the second micro-feature image dataset, and determining the multi-class micro-features corresponding to the second micro-feature image dataset; the multi-class micro-features include multiple single-class micro-features; and determining the multi-class micro-features contained in the second micro-feature image dataset as the label of the second training sample.

[0120] In some embodiments, the pre-trained regression prediction model is determined by: obtaining a training sample set; the training sample set includes: multiple training samples, prediction features of each training sample, and a target to be predicted for each training sample; the training samples include a first micro-feature image dataset and a second micro-feature image dataset; the prediction features of the training samples include the quantity corresponding to each micro-feature contained in the first micro-feature image dataset and the quantity corresponding to each micro-feature contained in the second micro-feature image dataset; the target to be predicted for the training samples includes ductile fracture and brittle fracture; based on the training sample set, the original extreme gradient boosting tree model is trained to obtain the pre-trained regression prediction model.

[0121] In some embodiments, the pre-trained regression prediction model satisfies the following formula:

[0122]

[0123] Where x is the dataset of microscopic feature images to be predicted; η represents the ductility and brittleness of metal fracture corresponding to the dataset of microscopic feature images to be predicted; T represents the total number of decision trees in the model; η represents the learning rate; q represents the ductility and brittleness of metal fracture. t (x) is the index of the leaf node to which the micro-feature image dataset to be predicted is mapped in the t-th tree; The value corresponding to the index of the leaf node.

[0124] In some embodiments, the microstructure includes at least one of dimples, intergranular structures, and cleavage.

[0125] In some embodiments, the processing unit 602 is specifically configured to: acquire a dataset of images of the microscopic features to be predicted on the fracture surface of a metal using a scanning electron microscope.

[0126] This application also provides a computer-readable storage medium, which includes computer-executable instructions. When the computer-executable instructions are executed on the computer, the computer performs the metal fracture brittleness identification method provided in the above embodiments.

[0127] This application also provides a computer program that can be directly loaded into a memory and contains software code. After being loaded and executed by a computer, the computer program can realize the metal fracture ductility and brittleness identification method provided in the above embodiments.

[0128] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this application can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer-readable storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer.

[0129] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and other division methods may exist in actual implementation. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms. Units described as separate components may or may not be physically separate; components shown as units may be one physical unit or multiple physical units, i.e., they may be located in one place or distributed in multiple different places. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0131] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, in essence, or the part that contributes to general technology, or all or part of the technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, ROM, RAM, magnetic disks, or optical disks.

[0132] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for identifying the ductile-brittle fracture properties of metals, characterized in that, include: Obtain a dataset of images of the microscopic features to be predicted on the fracture surface of a metal. The dataset of microscopic features to be predicted is input into a pre-trained microscopic feature recognition model to determine the quantity of each microscopic feature contained in the dataset of microscopic features to be predicted. The quantity corresponding to each micro-feature contained in the image dataset of micro-features to be predicted is input into a pre-trained regression prediction model to determine the ductility and brittleness of the metal fracture.

2. The method according to claim 1, characterized in that, The pre-trained microscopic feature recognition model is determined in the following way: Obtain a first training sample set; the first training sample set includes multiple first training samples and a label for each first training sample; the first training samples include a first microscopic feature image dataset; the labels of the first training samples include the microscopic features corresponding to the first microscopic feature image dataset; the first microscopic feature image dataset is an image set containing a single type of microscopic feature; Based on the first training sample set, the original deep residual network model is trained to obtain a single-class micro-feature classification model. Obtain a second training sample set, which includes multiple second training samples and a label for each second training sample; the second training sample set includes a second microscopic feature image dataset. The labels of the second training samples include multiple types of micro-features corresponding to the second micro-feature image dataset; the second micro-feature image dataset is an image set containing the multiple types of micro-features; Based on the second training sample set, the original deep residual network model is trained and combined with a single-class micro-feature classification model to obtain the pre-trained micro-feature recognition model.

3. The method according to claim 2, characterized in that, The label of each second training sample is determined in the following way: The second microscopic feature image dataset is input into the single-category microscopic feature classification model to determine the multiple single-category microscopic features contained in the second microscopic feature image dataset, thereby determining the multiple categories of microscopic features corresponding to the second microscopic feature image dataset. The various types of micro-features include the multiple single-type micro-features; The various micro-features contained in the second micro-feature image dataset are determined as the labels of the second training samples.

4. The method according to claim 1, characterized in that, The pre-trained regression prediction model is determined in the following way: Obtain a training sample set; the training sample set includes: multiple training samples, prediction features of each training sample, and a target to be predicted for each training sample; the training samples include a first microscopic feature image dataset and a second microscopic feature image dataset; the prediction features of the training samples include the quantity of each microscopic feature contained in the first microscopic feature image dataset and the quantity of each microscopic feature contained in the second microscopic feature image dataset; the targets to be predicted for the training samples include ductile fracture and brittle fracture; Based on the training sample set, the original extreme gradient boosting tree model is trained to obtain the pre-trained regression prediction model.

5. The method according to claim 4, characterized in that, The pre-trained regression prediction model satisfies the following formula: Where x is the dataset of microscopic feature images to be predicted; The ductility and brittleness of the metal fracture corresponding to the microscopic feature image dataset to be predicted; T is the total number of decision trees in the model; η is the learning rate; q t (x) is the index of the leaf node to which the dataset of microscopic features to be predicted is mapped in the t-th tree; ωq t (x) is the value corresponding to the index of the leaf node.

6. The method according to claim 1, characterized in that, The microscopic features include at least one of dimples, intergranularity, and cleavage.

7. The method according to claim 1, characterized in that, The dataset of images showing the microscopic features to be predicted on the fracture surface of the metal includes: A dataset of images of the microscopic features to be predicted on the fracture surface of the metal was obtained using a scanning electron microscope.

8. A device for identifying the brittle and ductile fracture properties of metals, characterized in that, include: Communication unit and processing unit; The communication unit is used to acquire a dataset of images of the microscopic features to be predicted on the fracture surface of the metal. The processing unit is used to input the micro-feature image dataset to be predicted into a pre-trained micro-feature recognition model to determine the quantity of each micro-feature contained in the micro-feature image dataset to be predicted. The processing unit is further configured to input the quantity corresponding to each micro-feature contained in the micro-feature image dataset to be predicted into a pre-trained regression prediction model to determine the ductility and brittleness of the metal fracture.

9. A device for identifying the brittle and tough fracture properties of metals, characterized in that, include: Including the processor and memory; The memory is used to store one or more programs, the one or more programs including computer-executable instructions. When the metal fracture brittleness identification device is running, the processor executes the computer-executable instructions stored in the memory to cause the metal fracture brittleness identification device to perform the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the computer execution instructions stored in the computer-readable storage medium are executed by the processor of the metal fracture brittleness identification device, the metal fracture brittleness identification device is capable of performing the method as described in any one of claims 1 to 7.

Images