Nickel-based superalloy internal organization structure characterization method, equipment, medium and product

Abstract

The invention discloses a nickel-based superalloy internal organization structure characterization method and device, a medium and a product, and relates to the field of nondestructive testing and visual characterization, and the method comprises the following steps: constructing an ultrasonic parameter evaluation model according to ultrasonic testing data and a microstructure image in each sample data; training a denoising diffusion generation network containing geometric constraints according to the geometric structure features corresponding to the sample data; the denoising diffusion generation network containing geometric constraints is constructed based on a geometric constraint sub-network and a U-Net prediction network; generating geometric structure characteristics corresponding to the ultrasonic detection data of the nickel-based superalloy to be characterized based on the ultrasonic parameter evaluation model; and according to the geometric structure characteristics, obtaining a corresponding microstructure structure image based on a denoising diffusion generation network containing geometric constraints. According to the method, the observability of the two-dimensional visual representation image in the nickel-based superalloy can be improved.

Description

Technical Field

[0001] The present application relates to the field of nondestructive testing and visual characterization, and in particular to a method, equipment, medium and product for characterizing the internal structure of a nickel-based high-temperature alloy. Background Art

[0002] Nickel-based superalloys have significant fatigue resistance, radiation resistance, oxidation resistance, corrosion resistance, and good processing and welding properties. They can maintain long-term organizational stability and are widely used in key components such as disks, blades, and casings in the aviation field. The material properties of nickel-based superalloys are closely related to the metallographic structure. In the actual use of superalloys, workpieces often serve for a long time in an environment of alternating stress and high temperature, which is prone to changes in local metallographic structure and causes safety problems. Therefore, it is very necessary to detect and accurately evaluate the metallographic structure of nickel-based superalloys.

[0003] Existing evaluation methods can be divided into two categories: destructive and non-destructive. Metallographic examination is a commonly used destructive test method that can accurately observe the microstructure of the material through a microscope. However, the use of random sampling in this technology may result in the inability to detect certain features. Non-destructive testing or evaluation can achieve 100% coverage of the analyzed material. Among them, ultrasonic non-destructive testing has the advantages of strong penetration, high sensitivity and resolution, and can be positioned and quantitatively detected. It has been widely used in the field of material manufacturing quality inspection. However, this detection method only uses numerical information to characterize and evaluate the internal structure of the alloy, losing the structural characteristics and position information between grains, and the numerical results lack observability.

[0004] Therefore, based on the above problems, there is an urgent need to provide a new method or system for characterizing the internal structure of nickel-based high-temperature alloys, so as to improve the visibility of the two-dimensional visualization characterization image inside the nickel-based high-temperature alloys. Summary of the invention

[0005] The purpose of this application is to provide a method, equipment, medium and product for characterizing the internal organizational structure of a nickel-based high-temperature alloy, which can make up for the deficiency of existing ultrasonic non-destructive testing that only uses data information to characterize the internal microstructure of the alloy, and can improve the visibility of the two-dimensional visualization characterization image inside the nickel-based high-temperature alloy.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a method for characterizing the internal structure of a nickel-based high-temperature alloy, the method comprising:

[0008] Acquire multiple sample data; the sample data includes: ultrasonic detection data of nickel-based high-temperature alloy and microstructure images observed under an optical microscope;

[0009] An ultrasonic parameter evaluation model is constructed based on the ultrasonic detection data and microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness and major-minor axis ratio;

[0010] According to the geometric structure features corresponding to each sample data, a denoising diffusion generation network with geometric constraints is trained; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network;

[0011] Obtain ultrasonic testing data of the nickel-based high-temperature alloy to be characterized;

[0012] Generate geometric structure features corresponding to ultrasonic testing data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model;

[0013] According to the geometric structure characteristics corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized, the corresponding microstructure image is obtained based on a denoising diffusion generation network with geometric constraints.

[0014] Optionally, the acquiring of multiple sample data specifically includes:

[0015] Use ultrasonic scanning equipment to obtain ultrasonic testing data of nickel-based high-temperature alloys;

[0016] Metallographic experiments are used to obtain microstructure images of alloy samples, and image processing algorithms are used to obtain internal geometric structural features;

[0017] Ultrasonic feature data is extracted from the ultrasonic detection data; the ultrasonic feature data includes: the longitudinal wave sound velocity, attenuation coefficient and frequency offset of the material.

[0018] Optionally, the image processing algorithm is used to obtain the internal geometric structure features, specifically including:

[0019] Using the formula Determine the grain size D;

[0020] Using the formula Determine the roundness M;

[0021] Using the formula Determine the major-minor axis ratio R;

[0022] Among them, Area is the grain area, Len is the grain length, a and b are the major axis and minor axis of the grain respectively, and π is the pi.

[0023] Optionally, constructing an ultrasound parameter evaluation model according to the ultrasound detection data and the microscopic tissue structure image in each sample data specifically includes:

[0024] Determine the corresponding geometric structure features according to the microscopic tissue structure image;

[0025] A quadratic polynomial mapping function is used to map the three types of geometric structure features into one-dimensional parameters;

[0026] A linear fitting function is used to fit the one-dimensional parameters and the corresponding ultrasonic detection data to construct the relationship between the geometric structure feature data and the ultrasonic detection data;

[0027] According to the relationship between geometric structure feature data and ultrasonic detection data, the optimization goal is to minimize the mean absolute error between one-dimensional parameters and actual geometric structure features. An evolutionary algorithm is used to solve the optimization problem and an ultrasonic parameter evaluation model is obtained.

[0028] Optionally, the training of a denoising diffusion generation network with geometric constraints according to the geometric structure features corresponding to each sample data specifically includes:

[0029] In the forward diffusion process of the denoising diffusion generation network with geometric constraints, Gaussian noise with a certain weight is gradually added to the input microstructure image, degenerating into a Gaussian noise image.

[0030] Use Gaussian noise image and corresponding geometric structure features as input, and perform encoding fusion through embedding function;

[0031] The fused vector is introduced into the middle feature layer of the U-Net network in the denoising diffusion generation network with geometric constraints through a linear superposition function;

[0032] The denoising diffusion generative network with geometric constraints is trained by calculating the MSE loss between the real denoised noise and the predicted denoised noise.

[0033] Optionally, a Markov process is used to gradually add Gaussian noise of a certain weight to the input microstructure image.

[0034] In a second aspect, the present application provides a nickel-based high-temperature alloy internal structure characterization device, the nickel-based high-temperature alloy internal structure characterization device comprising:

[0035] A data acquisition module is used to acquire a plurality of sample data; the sample data includes: ultrasonic detection data of nickel-based high-temperature alloy and microstructure images observed under an optical microscope;

[0036] An ultrasonic evaluation module is used to construct an ultrasonic parameter evaluation model according to the ultrasonic detection data and microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness and major-minor axis ratio;

[0037] A model training module, used for training a denoising diffusion generation network with geometric constraints according to the geometric structure features corresponding to each sample data; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network;

[0038] The microstructure generation module is used to obtain ultrasonic detection data of the nickel-based high-temperature alloy to be characterized; generate geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model; and obtain the corresponding microstructure image based on the denoising diffusion generation network containing geometric constraints according to the geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized.

[0039] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for characterizing the internal microstructure of a nickel-based high-temperature alloy.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for characterizing the internal microstructure of a nickel-based high-temperature alloy.

[0041] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for characterizing the internal organizational structure of a nickel-based high-temperature alloy.

[0042] According to the specific embodiments provided in this application, this application has the following technical effects:

[0043] The present application provides a method, device, medium and product for characterizing the internal microstructure of a nickel-based high-temperature alloy, by obtaining ultrasonic detection data of a plurality of nickel-based high-temperature alloys and microstructure images observed under an optical microscope; using an ultrasonic parameter evaluation model to obtain the geometric structural characteristics of the microstructure inside the alloy; using a denoising diffusion generation model containing a geometric constraint subnetwork to achieve alloy microstructure characterization based on ultrasonic detection data; the present application can achieve non-destructive characterization of the microstructure inside the alloy by ultrasonic detection, making up for the deficiency of existing ultrasonic non-destructive testing that only uses data information to characterize the microstructure inside the alloy, and obtaining a two-dimensional visual characterization image of the alloy inside through the ultrasonic detection signal of the alloy is more observable. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 This is a schematic diagram of a method for characterizing the internal structure of a nickel-based high-temperature alloy in one embodiment of the present application;

[0046] Figure 2 This is a flow chart of characterizing the internal structure of the GH4169 alloy based on ultrasonic testing data in one embodiment of the present application;

[0047] Figure 3 This is a schematic diagram of the internal structure characterization of the GH4169 alloy based on ultrasonic testing data in one embodiment of the present application;

[0048] Figure 4 The ultrasonic signals detected by ultrasonic testing equipment, the microstructure diagram collected by optical microscope, and the geometric characteristic data of the microstructure of GH4169 after two different heat treatment processes. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0051] In an exemplary embodiment, Figure 1 As shown, a method for characterizing the internal structure of a nickel-based high-temperature alloy is provided, the method comprising the following S101 to S106. Wherein:

[0052] S101, acquiring a plurality of sample data; the sample data comprising: ultrasonic testing data of a nickel-based high-temperature alloy and a microstructure image observed under an optical microscope; the microstructure image observed under an optical microscope comprises: microstructure images of all ultrasonic testing areas;

[0053] S101 specifically includes:

[0054] S11, using ultrasonic scanning equipment to obtain ultrasonic testing data of nickel-based high-temperature alloy;

[0055] S12, using metallographic experiments to obtain microstructure images of alloy samples, and using image processing algorithms to obtain internal geometric structure features;

[0056] Using the formula Determine the grain size D;

[0057] Using the formula Determine the roundness M;

[0058] Using the formula Determine the major-minor axis ratio R;

[0059] Among them, Area is the grain area, Len is the grain length, a and b are the major axis and minor axis of the grain respectively, and π is the pi.

[0060] S13, extracting ultrasonic feature data from the ultrasonic detection data; the ultrasonic feature data Y includes: the longitudinal wave sound velocity C of the material L , attenuation coefficient α and frequency offset Δf.

[0061] S102, constructing an ultrasonic parameter evaluation model according to the ultrasonic detection data and the microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness, and major-minor axis ratio;

[0062] S102 specifically includes:

[0063] S21, determining corresponding geometric structure features according to the microscopic tissue structure image;

[0064] S22, using a quadratic polynomial mapping function to map the three types of geometric structure features into one-dimensional parameters, to obtain the ultrasonic evaluation data X * (Grain size Roundness Axis ratio );

[0065] S23, fitting the one-dimensional parameters and the corresponding ultrasonic detection data using a linear fitting function to construct a relationship between the geometric structure feature data and the ultrasonic detection data;

[0066] S24, based on the relationship between geometric structure feature data and ultrasonic detection data, taking the minimum mean absolute error between one-dimensional parameters and actual geometric structure features as the optimization goal, using evolutionary algorithm to solve the optimization problem and obtain the ultrasonic parameter evaluation model; optimization goal Where m is the total number of samples;

[0067] The ultrasound parameter evaluation model specifically includes:

[0068] Mapping function:

[0069] Fitting function:

[0070] Among them, λ is the mapping parameter, ξ is the fitting parameter;

[0071] S103, training a denoising diffusion generation network with geometric constraints according to the geometric structure features corresponding to each sample data; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network;

[0072] S103 specifically includes:

[0073] S31, in the forward diffusion process of the denoising diffusion generation network with geometric constraints, Gaussian noise with a certain weight is gradually added to the input microstructure image, degenerating into a Gaussian noise image;

[0074] The Markov process is used to add noise to the microstructure image. The noisy image at time t where x t-1 is the noisy image at the previous moment, is the weight of the noisy image at the previous moment to the noisy image at the current moment;

[0075] The weight is gradually increased. It gradually increases with the increase of time step t, which is conducive to speeding up the forward noise addition process.

[0076] S32, using Gaussian noise image and corresponding geometric structure features as input, and performing encoding fusion through embedding function;

[0077] S33, introducing the fused vector into the middle feature layer of the U-Net network in the denoising diffusion generation network with geometric constraints through a linear superposition function;

[0078] The U-Net network is used in the reverse diffusion process, and can predict the noise data in the forward diffusion process, and then restore the Gaussian noise image that has lost its structural features to a microscopic tissue structure image; the geometric constraint subnetwork takes the geometric structure data of the microscopic tissue structure as input, and regulates the U-Net network output through encoding fusion.

[0079] Image processing is used to obtain the geometric structural features of the original training microstructure image and the restored image similar to the original training microstructure image, including grain size, roundness and major-minor axis ratio, and then the mean square error between the above geometric features is obtained. The calculation method is as follows:

[0080] The U-Net network is used to predict the noise ε in the forward noise addition process. θ (x t ,t), use Bayesian theorem to restore to a less noisy version;

[0081] The coding fusion method is used to fuse the geometric constraint data into the U-Net network to guide network prediction.

[0082] S34, by calculating the MSE loss between the real noise added and the predicted noise added, a denoising diffusion generation network with geometric constraints is trained.

[0083] S104, obtaining ultrasonic testing data of the nickel-based high-temperature alloy to be characterized;

[0084] S105, generating geometric structure features corresponding to ultrasonic detection data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model;

[0085] S106, according to the geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized, a corresponding microstructure image is obtained based on a denoising diffusion generation network containing geometric constraints.

[0086] In practical application, the operation steps of this application are as follows:

[0087] The ultrasonic signal data and microstructure images of GH4169 alloy samples were obtained, and the data were screened and evaluated.

[0088] Different heat treatment processes were adopted for the nickel-based high-temperature alloy GH4169 to obtain GH4169 alloy samples with different internal grain distribution states.

[0089] Ultrasonic scanning equipment was used to perform ultrasonic testing on the GH4169 alloy samples after heat treatment to obtain ultrasonic testing signal data.

[0090] The heat-treated GH4169 alloy specimens were pre-tested, mechanically ground, polished, corroded, and observed under a microscope to collect microstructural images of the GH4169 alloy specimens.

[0091] The quality of the acquired GH4169 microstructure images was evaluated, and the microstructure images with poor performance were eliminated.

[0092] Ultrasonic feature data is extracted from the ultrasonic detection signal; specifically, the longitudinal wave speed CL , the attenuation coefficient α and the frequency offset Δf are obtained by the calculation formula, where: Δf=|f1-f2|, where h is the sample thickness, H1 and H2 are the highest peak values ​​of the primary and secondary bottom waves respectively, and f1 and f2 are the frequencies of the primary and secondary bottom waves respectively.

[0093] The geometric features of the microstructure image are extracted; specifically, the grain size D, roundness M and major-minor axis ratio R are used. Where Area is the grain area, Len is the grain length, a and b are the major axis and minor axis of the grain, respectively.

[0094] An ultrasound parameter evaluation model is constructed based on the acquired ultrasound feature data and the geometric feature data of the microscopic tissue structure image.

[0095] The grain size D, roundness M and major-minor axis ratio R are mapped to a one-dimensional parameter Z through a quadratic polynomial mapping function.

[0096] By linear fitting function The one-dimensional parameters are fitted with the ultrasonic parameter data to construct the relationship between the geometric structure feature data and the ultrasonic feature data.

[0097] In the fitting process, the parameters of the fitting and mapping functions are obtained by constructing a minimization optimization objective and using an evolutionary algorithm to solve the optimization problem.

[0098] In the forward diffusion process of the denoising diffusion generation network with geometric constraints, Gaussian noise with a certain weight is gradually added to the input microstructure image, so that the internal structural features of the image are gradually destroyed and degenerated into a Gaussian noise image.

[0099] The noise adding process is a Markov process, and the state at the current moment is determined by the state at the previous moment. where x t-1 is the noisy image at the previous moment, is the weight of the noisy image at the previous moment to the noisy image at the current moment.

[0100] Among them, as time t gradually increases, the weight Gradually increase.

[0101] The U-Net network guided by the training geometric constraint subnet realizes the reverse diffusion generation controlled by grain size, roundness and aspect ratio, and then restores the Gaussian noise image that has lost its structural features into a microstructure image.

[0102] The geometric sub-network uses the grain size D, roundness M and major-minor axis ratio R of the GH4169 alloy microstructure image and the current diffusion time step t in the back diffusion process as input, encodes and fuses them through the embedding function, and then introduces the fused vector into the intermediate feature layer of the U-Net network through the linear superposition function.

[0103] The U-Net network with a constrained sub-network is used to predict the denoising noise of the forward process. By calculating the MSE loss between the real denoising noise and the predicted denoising noise, the training of the denoising diffusion generation model with a constrained sub-network is realized.

[0104] The geometric structure features obtained by the ultrasonic parameter evaluation model are input into the trained denoising diffusion generation network with geometric constraints to achieve non-destructive characterization of the microstructure inside the alloy by ultrasonic testing.

[0105] For the GH4169 alloy to be tested, ultrasonic scanning equipment is used to perform ultrasonic data detection on the alloy sample to be tested.

[0106] The ultrasonic parameter evaluation model is used to convert the detected ultrasonic data into corresponding geometric feature data.

[0107] The geometric feature data obtained by conversion is used as the input of the trained diffusion generation network with geometric constraints to obtain the internal microstructure image of the alloy sample to be tested.

[0108] As a specific example, Figure 2 and Figure 3 As shown, the specific implementation process is as follows:

[0109] K1: Acquisition and effectiveness evaluation of ultrasound signal data and microscopic tissue structure images, including the following steps:

[0110] K11: Different heat treatment processes are used on the nickel-based high-temperature alloy GH4169 to obtain GH4169 alloy samples with different internal grain distribution states.

[0111] K12: Use ultrasonic scanning equipment to perform ultrasonic testing on the heat-treated GH4169 alloy sample to obtain ultrasonic testing signal data.

[0112] K13: The heat-treated GH4169 alloy specimen was pre-tested, mechanically ground, polished, corroded, and observed under a microscope to collect the microstructure image of the GH4169 alloy specimen.

[0113] K14: Evaluate the quality of the acquired GH4169 microstructure images and remove the microstructure images with poor performance.

[0114] K2: Ultrasonic feature data extraction of ultrasonic detection signals, including the following steps:

[0115] K21: Return the reference value of the ultrasonic scanning signal to zero. Specifically, adjust the amplitude of the ultrasonic signal so that the signal horizontal baseline amplitude is close to zero.

[0116] K22: Record the highest peaks H1 and H2 of the primary and secondary bottom waves of the sample thickness h, as well as the propagation time T from the primary bottom wave to the secondary bottom wave.

[0117] K23: Using the longitudinal wave speed C L , the attenuation coefficient α and the frequency offset Δf are obtained by the calculation formula, where: Δf=|f1-f2|, where f1 and f2 are the frequencies of the primary bottom wave and the secondary bottom wave respectively.

[0118] K3: Geometric feature extraction of microstructure images, including the following steps:

[0119] K31: The contour detection method is used to segment the grains inside the microstructure of the GH4169 alloy, and the grain area Area, grain length Len, and the major and minor axes a and b of the grains are calculated.

[0120] K32: Use grain size D, roundness M and major-minor axis ratio R. Among them,

[0121] K4: Construct an ultrasound parameter evaluation model, including the following steps:

[0122] K41: The grain size D, roundness M and major-minor axis ratio R are mapped into one-dimensional data Z through a quadratic polynomial mapping function.

[0123] K42: By linear fitting function The one-dimensional data is fitted with the ultrasonic parameter data to construct the relationship between the geometric structure feature data and the ultrasonic feature data.

[0124] K43: During the fitting process, the parameters of the fitting and mapping functions are obtained by constructing a minimization optimization objective and using an evolutionary algorithm to solve the optimization problem.

[0125] K5: Forward diffusion and noise addition of microstructure image, including the following steps:

[0126] K51: The noise addition process is a Markov process. The state at the current moment is determined by the state at the previous moment. The noisy image at time t where x t-1 is the noisy image at the previous moment, is the weight of the noisy image at the previous moment to the noisy image at the current moment.

[0127] K52: As time t increases, the weight Gradually increase.

[0128] K6: Backward diffusion generation of the subnetwork with geometric constraints, including the following steps:

[0129] K61: The geometric subnetwork uses the grain size D, roundness M and major-minor axis ratio R of the microstructure image of GH4169 alloy and the current diffusion time step t in the back diffusion process as input, encodes and fuses them through the embedding function, and then introduces the fused vector into the intermediate feature layer of the U-Net network through the linear superposition function.

[0130] K62: The U-Net network with a constrained sub-network is used to predict the denoising noise of the forward process. By calculating the MSE loss between the real denoising noise and the predicted denoising noise, the denoising diffusion generation model with a constrained sub-network is trained.

[0131] K7: Image characterization of the microstructure of GH4169 alloy guided by ultrasonic testing data, including the following steps:

[0132] K71: For the GH4169 alloy to be tested, ultrasonic data detection is performed on the alloy sample using ultrasonic scanning equipment.

[0133] K72: The ultrasonic parameter evaluation model is used to convert the detected ultrasonic data into corresponding geometric feature data.

[0134] K73: The converted geometric feature data is used as the input of the trained geometrically constrained denoising diffusion generation network to obtain the internal microstructure image of the alloy sample to be tested.

[0135] Based on the same inventive concept, the embodiment of the present application also provides a nickel-based high-temperature alloy internal structure characterization device for implementing the nickel-based high-temperature alloy internal structure characterization method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more nickel-based high-temperature alloy internal structure characterization device embodiments provided below can refer to the limitations of the nickel-based high-temperature alloy internal structure characterization method above, and will not be repeated here.

[0136] In an exemplary embodiment, a nickel-based high-temperature alloy internal structure characterization device is provided, comprising:

[0137] A data acquisition module is used to acquire a plurality of sample data; the sample data includes: ultrasonic detection data of nickel-based high-temperature alloy and microstructure images observed under an optical microscope;

[0138] An ultrasonic evaluation module is used to construct an ultrasonic parameter evaluation model according to the ultrasonic detection data and microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness and major-minor axis ratio;

[0139] A model training module, used for training a denoising diffusion generation network with geometric constraints according to the geometric structure features corresponding to each sample data; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network;

[0140] The microstructure generation module is used to obtain ultrasonic detection data of the nickel-based high-temperature alloy to be characterized; generate geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model; and obtain the corresponding microstructure image based on the denoising diffusion generation network containing geometric constraints according to the geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized.

[0141] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for characterizing the internal structure of a nickel-based high-temperature alloy is implemented.

[0142] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0143] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0144] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0145] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0146] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0147] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0148] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0149] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for characterizing the internal structure of a nickel-based high-temperature alloy, characterized in that: The method for characterizing the internal structure of the nickel-based high-temperature alloy comprises: Acquire multiple sample data; the sample data includes: ultrasonic detection data of nickel-based high-temperature alloy and microstructure images observed under an optical microscope; An ultrasonic parameter evaluation model is constructed based on the ultrasonic detection data and microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness and major-minor axis ratio; According to the geometric structure features corresponding to each sample data, a denoising diffusion generation network with geometric constraints is trained; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network; Obtain ultrasonic testing data of the nickel-based high-temperature alloy to be characterized; Generate geometric structure features corresponding to ultrasonic testing data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model; According to the geometric structure characteristics corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized, the corresponding microstructure image is obtained based on a denoising diffusion generation network with geometric constraints.

2. The method for characterizing the internal structure of a nickel-based high-temperature alloy according to claim 1, characterized in that: The obtaining of multiple sample data specifically includes: Use ultrasonic scanning equipment to obtain ultrasonic testing data of nickel-based high-temperature alloys; Metallographic experiments are used to obtain microstructure images of alloy samples, and image processing algorithms are used to obtain internal geometric structural features; Ultrasonic feature data is extracted from the ultrasonic detection data; the ultrasonic feature data includes: the longitudinal wave sound velocity, attenuation coefficient and frequency offset of the material.

3. The method for characterizing the internal structure of a nickel-based high-temperature alloy according to claim 2, characterized in that: The image processing algorithm is used to obtain the internal geometric structure features, specifically including: Using the formula Determine the grain size D; Using the formula Determine the roundness M; Using the formula Determine the major-minor axis ratio R; Among them, Area is the grain area, Len is the grain length, a and b are the major axis and minor axis of the grain respectively, and π is the pi.

4. The method for characterizing the internal structure of a nickel-based high-temperature alloy according to claim 1, characterized in that: The ultrasonic parameter evaluation model is constructed according to the ultrasonic detection data and the microscopic tissue structure image in each sample data, specifically including: Determine the corresponding geometric structure features according to the microscopic tissue structure image; A quadratic polynomial mapping function is used to map the three types of geometric structure features into one-dimensional parameters; A linear fitting function is used to fit the one-dimensional parameters and the corresponding ultrasonic detection data to construct the relationship between the geometric structure feature data and the ultrasonic detection data; According to the relationship between geometric structure feature data and ultrasonic detection data, the optimization goal is to minimize the mean absolute error between one-dimensional parameters and actual geometric structure features. An evolutionary algorithm is used to solve the optimization problem and an ultrasonic parameter evaluation model is obtained.

5. The method for characterizing the internal structure of a nickel-based high-temperature alloy according to claim 1, characterized in that: The method of training a denoising diffusion generation network containing geometric constraints according to the geometric structure features corresponding to each sample data specifically includes: In the forward diffusion process of the denoising diffusion generation network with geometric constraints, Gaussian noise with a certain weight is gradually added to the input microstructure image, degenerating into a Gaussian noise image. Use Gaussian noise image and corresponding geometric structure features as input, and perform encoding fusion through embedding function; The fused vector is introduced into the middle feature layer of the U-Net network in the denoising diffusion generation network with geometric constraints through a linear superposition function; The denoising diffusion generative network with geometric constraints is trained by calculating the MSE loss between the real denoised noise and the predicted denoised noise.

6. The method for characterizing the internal structure of a nickel-based high-temperature alloy according to claim 5, characterized in that: The Markov process is used to gradually add Gaussian noise with a certain weight to the input microstructure image.

7. A nickel-based high-temperature alloy internal structure characterization device, characterized in that: The nickel-based high-temperature alloy internal structure characterization equipment includes: A data acquisition module is used to acquire a plurality of sample data; the sample data includes: ultrasonic detection data of nickel-based high-temperature alloy and microstructure images observed under an optical microscope; An ultrasonic evaluation module is used to construct an ultrasonic parameter evaluation model according to the ultrasonic detection data and microstructure image in each sample data; the ultrasonic parameter evaluation model is used to output the geometric structure characteristics of the internal microstructure of the nickel-based high-temperature alloy; the geometric structure characteristics include: grain size, roundness and major-minor axis ratio; A model training module, used for training a denoising diffusion generation network with geometric constraints according to the geometric structure features corresponding to each sample data; the denoising diffusion generation network with geometric constraints is constructed based on the geometric constraint sub-network and the U-Net prediction network; The microstructure generation module is used to obtain ultrasonic detection data of the nickel-based high-temperature alloy to be characterized; generate geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized based on the ultrasonic parameter evaluation model; and obtain the corresponding microstructure image based on the denoising diffusion generation network containing geometric constraints according to the geometric structure features corresponding to the ultrasonic detection data of the nickel-based high-temperature alloy to be characterized.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for characterizing the internal microstructure of a nickel-based high-temperature alloy according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for characterizing the internal microstructure of a nickel-based high-temperature alloy according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for characterizing the internal microstructure of a nickel-based high-temperature alloy according to any one of claims 1 to 6 is implemented.

Images