Annealing temperature-based titanium alloy forge piece structure property improving method and system

Abstract

The invention relates to the technical field of alloy forging, in particular to a titanium alloy forge piece structure performance improving method and system based on the annealing temperature. Performing stress distribution identification operation on the forging process information to obtain a stress distribution matrix, and performing weighted calculation on the initial two-phase distribution information according to the stress distribution matrix to obtain two-phase distribution information of the forge piece; performing feature extraction operation on the two-phase distribution information of the forge piece to obtain a two-phase distribution feature set, and performing local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and predict primary recovery two-phase distribution; according to the performance weight distribution and the initial two-phase distribution information, performing global annealing prediction operation on the predicted primary recovery two-phase distribution to obtain a global annealing step; and carrying out annealing operation according to the local annealing step and the global annealing step. According to the forging application scene, the titanium alloy structure performance can be improved to the maximum extent through annealing treatment.

Description

Technical Field

[0001] The present invention relates to the technical field of alloy forging, and in particular to a method and system for improving the microstructure and properties of titanium alloy forgings based on annealing temperature. Background Art

[0002] Most titanium alloy materials are distributed Related Phase. Among them, The phase is a hexagonal close-packed (HCP) structure, which remains stable at room temperature and in a higher temperature range, providing good creep resistance, excellent corrosion resistance and high strength. The phase is a body-centered cubic (BCC) structure, which has good plasticity and can remain stable at high temperatures. In order to ensure that the structural properties of titanium alloys meet the requirements of application scenarios, most of the production process uses heat treatment technologies such as annealing to adjust the titanium alloy. Related The distribution of phases can be adjusted to fine-tune the microstructure and properties of titanium alloys.

[0003] However, today's titanium alloy annealing process is generally based on standardized titanium alloy annealing standards. Technicians subjectively select appropriate standards for annealing and then conduct performance testing. If the test meets the standards, the process is completed, but it is unknown whether the titanium alloy's performance has reached the optimal state required. Summary of the Invention

[0004] The present invention provides a method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature, the main purpose of which is to maximize the microstructure and properties of titanium alloy through annealing treatment according to the application scenarios of forgings.

[0005] To achieve the above-mentioned purpose, the present invention provides a method for improving the organizational properties of titanium alloy forgings based on annealing temperature, comprising: obtaining a pre-constructed application scenario of a titanium alloy forging, and weighting a preset performance indicator set according to the application scenario to obtain a performance weight distribution; obtaining initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtaining forging process information of the titanium alloy forging, and obtaining a pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network; using the two-phase distribution recognition network, performing a stress distribution recognition operation on the forging process information to obtain a stress distribution matrix, and according to the The stress distribution matrix performs weighted calculation on the initial two-phase distribution information to obtain the two-phase distribution information of the forging; the annealing recognition network is used to perform feature extraction on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and a local annealing prediction operation based on two-phase recovery is performed according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution; according to the performance weight distribution and the initial two-phase distribution information, a global annealing prediction operation is performed on the predicted primary recovery two-phase distribution to obtain a global annealing step; according to the local annealing step and the global annealing step, a secondary annealing operation is performed on the titanium alloy forging to obtain a performance-enhanced forging.

[0006] Optionally, the weight configuration of the preset performance indicator set according to the application scenario to obtain a performance weight distribution includes: performing text segmentation processing on the application scenario to obtain a text segmentation set, and performing a sequential quantization operation on the text segmentation set to obtain a text quantization set; performing a self-attention enhancement operation on the text quantization set to obtain a text enhancement vector sequence; performing a semantic recognition operation based on the preset performance indicator set on the text enhancement vector sequence to obtain a semantic demand distribution, and performing keyword recognition based on the performance indicator set on the text enhancement vector sequence to obtain a demand word set; performing clustering statistics on the demand word set to obtain a primary demand distribution, and performing weighted calculation on the semantic demand distribution based on the primary demand distribution to obtain a performance weight distribution.

[0007] Optionally, obtaining the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging includes: obtaining a microstructure image of the titanium alloy material of the titanium alloy forging, and performing Gaussian filtering on the microstructure image to obtain a noise-reduced tissue image; performing a feature extraction operation on the noise-reduced tissue image to obtain an image feature set, and performing a two-phase-based classification and judgment operation on the image feature set to obtain an alpha phase set and a beta phase set; performing a position frame selection and marking operation on the alpha phase set and the beta phase set to obtain initial two-phase distribution information.

[0008] Optionally, obtaining a pre-trained forging annealing recognition model includes: obtaining a historical annealing two-phase change law and a training sample set, obtaining an initialized two-phase distribution recognition network and an initialized annealing recognition network, wherein the initialized annealing recognition network includes an annealing temperature-performance change recognition network and a performance score maximization network; obtaining an annealing temperature law change model, and using the annealing temperature law change model to perform transfer learning on the annealing temperature-performance change recognition network to obtain an updated annealing temperature-performance change recognition network; using the historical annealing two-phase change law to train the updated annealing temperature-performance change recognition network and the performance score maximization network to obtain a trained annealing recognition network; using the training sample set to train the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network; and using the trained two-phase distribution recognition network and the annealing recognition network to construct a forging annealing recognition model.

[0009] Optionally, the use of the training sample set to train the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network includes: extracting a target sample from the training sample set in sequence, wherein the target sample includes the historical titanium alloy initial two-phase distribution, the historical titanium alloy forging process and the historical titanium alloy two-phase detection after forging; using the initialized two-phase distribution recognition network to extract the forging features of the historical titanium alloy forging process to obtain a forging feature set; according to the forging feature set, obtaining a predicted stress distribution, and according to the predicted stress distribution and the historical titanium alloy initial two-phase distribution, predicting a predicted two-phase distribution; calculating the predicted two-phase distribution. After forging with the historical titanium alloy, the loss value of the two phases is detected, and the loss value is minimized to obtain the network model parameters when the loss value is minimum, and the inverse network parameters of the initialized two-phase distribution recognition network are updated according to the network model parameters to obtain an updated two-phase distribution recognition network; it is determined whether the training sample set is a null value; when the training sample set is not a null value, returning to the above-mentioned step of extracting a target sample from the training sample set in sequence; when the training sample set is a null value, the updated two-phase distribution recognition network corresponding to each target sample in the training sample set is calculated based on the mean of the network model parameters to obtain a trained two-phase distribution recognition network.

[0010] Optionally, the two-phase distribution identification network is expressed as: , , , where represents the stress distribution matrix, represents the stress identification process, represents the distribution of alloy stress characteristics, Indicates the forging angle of the forging. Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, and Represents the influence coefficient of stress on the two phases.

[0011] Optionally, the method of using the annealing recognition network to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and performing a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and predict a primary recovery two-phase distribution includes: obtaining an objective function in the annealing recognition network, wherein the objective function is expressed as: , where represents the target two-phase distribution, represents the set of annealing schemes, Indicates the permutation and combination of the steps in the annealing scheme set. Representing the existing two-phase distribution; performing a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set; performing a classification operation based on the forging position on the two-phase distribution feature set to obtain the two-phase distribution characteristics of the forging area and the two-phase distribution characteristics of the normal area; according to the objective function, configuring the two-phase distribution characteristics of the forging area as the existing two-phase distribution, configuring the two-phase distribution characteristics of the normal area as the target two-phase distribution, performing a local annealing step identification operation to obtain a local annealing step; using the annealing identification network, according to the local annealing step, obtaining a predicted primary recovery two-phase distribution.

[0012] Optionally, the global annealing prediction operation is performed on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step, including: using the annealing identification network to perform a two-phase distribution prediction operation to maximize the performance score of the initial two-phase distribution information according to the performance weight distribution to obtain an optimal two-phase distribution; performing two-phase distribution gap identification on the optimal two-phase distribution and the predicted primary recovery two-phase distribution to obtain a two-phase distribution difference; obtaining an annealing step set and an annealing temperature range, and randomly combining annealing schemes that meet the two-phase distribution difference according to the annealing step set and the annealing temperature range to obtain an annealing scheme set; according to a preset annealing cost calculation strategy, screening the annealing scheme with the lowest annealing cost in the annealing scheme set to obtain an optimal annealing scheme, and obtaining a global annealing step corresponding to the optimal annealing scheme.

[0013] Optionally, after performing a secondary annealing operation on the titanium alloy forging according to the local annealing step and the global annealing step to obtain a performance-enhanced forging, the method further includes: performing parameter offset on the local annealing step and the global annealing step according to a preset offset to obtain an offset annealing step set; performing an annealing operation on the titanium alloy forging according to the offset annealing step set to obtain an offset annealing forging set, and performing physical testing on the offset annealing forging set to obtain an offset forging test result set; performing physical testing on the performance-enhanced forging to obtain a forging test result; calculating the matching degree between the forging test result and the offset forging test result set and the application scenario respectively to obtain a matching score set, and screening the matching score with the largest value in the matching score set to obtain the best matching score; judging whether the best matching score corresponds to the performance-enhanced forging; when the best matching score does not correspond to the performance-enhanced forging, obtaining the offset annealing step corresponding to the best matching score, and training the forging annealing recognition model according to the offset annealing step to obtain an optimized forging annealing recognition model.

[0014] To achieve the above-mentioned purpose, the present invention also provides a titanium alloy forging microstructure performance improvement system based on annealing temperature, comprising: a forging performance target acquisition module, used to obtain a pre-built application scenario of the titanium alloy forging, and weight-configure a preset performance indicator set according to the application scenario to obtain a performance weight distribution; an information-model acquisition module, used to obtain the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtain the forging process information of the titanium alloy forging, and obtain a pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network; an annealing step recognition module, used to use the two-phase distribution recognition network to perform a stress distribution recognition operation on the forging process information to obtain to the stress distribution matrix, and perform weighted calculation on the initial two-phase distribution information according to the stress distribution matrix to obtain the two-phase distribution information of the forging, and use the annealing recognition network to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and perform a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution, and perform a global annealing prediction operation on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step; a step execution module is used to perform a secondary annealing operation on the titanium alloy forging according to the local annealing step and the global annealing step to obtain a performance-enhanced forging.

[0015] In order to solve the above problems, the present invention also provides an electronic device, which includes: a memory storing at least one instruction; and a processor executing the instructions stored in the memory to implement the above-mentioned method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature.

[0016] In order to solve the above problems, the present invention also provides a computer-readable storage medium, which stores at least one instruction, and the at least one instruction is executed by a processor in an electronic device to implement the above-mentioned method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature.

[0017] The present invention solves the problem described in the background technology. The present invention first obtains the performance weight distribution of the target according to the application scenario. Through the performance weight distribution, we can know the two-phase ratio that we finally want to adjust the titanium alloy; then obtain the forging process information and initial two-phase distribution information of the titanium alloy forging. Through the forging process information, we can know the deformation of the titanium alloy material, thereby inferring the stress at the titanium alloy forging place, and through machine learning, we can learn the mapping relationship between stress and two-phase distribution changes, so as to predict the two-phase distribution information of the forging after forging; then, the present invention first takes the unforged normal area as the target, and through local annealing, restores the two phases of the forging area as much as possible, and then through global annealing, further reduces the difference between the forging area and the normal area, and can change the two-phase distribution of the titanium alloy during the global annealing process, thereby achieving overall organizational performance improvement. Therefore, the present invention can maximize the organizational performance of the titanium alloy through annealing treatment according to the forging application scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic flow chart of a method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to one embodiment of the present invention; Figure 2 This is a functional module diagram of a system for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to an embodiment of the present invention; Figure 3 A schematic structural diagram of an electronic device for implementing the method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to an embodiment of the present invention.

[0019] Explanation of the accompanying drawings: 1. electronic device; 10. processor; 11. memory; 12. bus.

[0020] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0021] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0022] The embodiment of the present application provides a method for improving the microstructure and performance of titanium alloy forgings based on annealing temperature. The execution subject of the method for improving the microstructure and performance of titanium alloy forgings based on annealing temperature includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for improving the microstructure and performance of titanium alloy forgings based on annealing temperature can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0023] Reference Figure 1 FIG2 is a flow chart of a method for improving the microstructure and performance of titanium alloy forgings based on annealing temperature according to an embodiment of the present invention. In this embodiment, the method for improving the microstructure and performance of titanium alloy forgings based on annealing temperature includes: S1. Obtaining a pre-established application scenario for titanium alloy forgings, and weighting a preset set of performance indicators according to the application scenario to obtain a performance weight distribution.

[0024] Among them, the application scenarios of the titanium alloy forgings, for example: the product is used on the engine casing, which requires high temperature stability, corrosion resistance, structural stability, etc....

[0025] The performance weight distribution refers to the weight distribution of indicators such as creep resistance, corrosion resistance, stable strength, toughness, and plasticity.

[0026] In detail, in an embodiment of the present invention, the weight configuration of the preset performance indicator set according to the application scenario to obtain a performance weight distribution includes: performing text segmentation processing on the application scenario to obtain a text segmentation set, and performing a sequential quantization operation on the text segmentation set to obtain a text quantization set; performing a self-attention enhancement operation on the text quantization set to obtain a text enhancement vector sequence; performing a semantic recognition operation based on a preset performance indicator set on the text enhancement vector sequence to obtain a semantic demand distribution, and performing keyword recognition based on the performance indicator set on the text enhancement vector sequence to obtain a demand word set; performing clustering statistics on the demand word set to obtain a primary demand distribution, and performing weighted calculation on the semantic demand distribution based on the primary demand distribution to obtain a performance weight distribution.

[0027] Among them, the performance index set includes indicators such as hardness, plasticity, toughness, and corrosion resistance.

[0028] Specifically, in an embodiment of the present invention, word segmentation and quantization operations can be performed using a Chinese word segmentation tool, but the quantization process is disordered and the order of the word segmentation characters needs to be marked to obtain a text quantization set. Then, in an embodiment of the present invention, the text quantization set is contextually enhanced based on the self-attention mechanism to obtain a text enhancement vector sequence. In an embodiment of the present invention, the text of the application scenario may accurately record the requirements for indicators such as hardness, plasticity, and toughness, or may not express specific proportions and only describe them through functional requirements.

[0029] Therefore, in the embodiment of the present invention, semantic recognition and keyword recognition are used to weight each performance indicator, and a semantic demand distribution and a demand word set are obtained. Among them, the demand words identified by keywords can be counted by clustering algorithm, and the more times a word appears, the more important it is.

[0030] The present invention then balances the two through weighted calculation to obtain a performance weight distribution. In addition, in embodiments of the present invention, the number of indicators in the performance indicator set may be large, while the performance indicators involved in the document may be small. The present invention can configure the weights of indicators not involved in the document for the application scenario by configuring the minimum standard of the titanium alloy material.

[0031] S2. Obtain initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtain forging process information of the titanium alloy forging, and obtain a pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network.

[0032] The initial two-phase distribution information of the titanium alloy material refers to the two-phase distribution information of the titanium alloy forging in the state of titanium alloy rod or titanium alloy plate. Related Phase. Among them, The phase is a hexagonal close-packed (HCP) structure, which remains stable at room temperature and in a higher temperature range, providing good creep resistance, excellent corrosion resistance and high strength. The phase has a body-centered cubic (BCC) structure, has good plasticity, and can remain stable at high temperatures.

[0033] The forging process information refers to information such as the number and angle of bending, folding, and forging of the titanium alloy material.

[0034] The forging annealing identification model refers to a model that uses annealing methods to make the microstructure and performance of forgings more suitable for the application scenarios of forgings, thereby improving the quality of forgings and extending the service life of forgings.

[0035] In detail, in an embodiment of the present invention, the obtaining of the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging includes: obtaining a microstructure image of the titanium alloy material of the titanium alloy forging, and performing Gaussian filtering on the microstructure image to obtain a noise-reduced tissue image; performing a feature extraction operation on the noise-reduced tissue image to obtain an image feature set, and performing a two-phase-based classification and judgment operation on the image feature set to obtain an alpha phase set and a beta phase set; performing a position frame selection and marking operation on the alpha phase set and the beta phase set to obtain the initial two-phase distribution information.

[0036] The microstructure image refers to an image that can observe the internal structure of the metal.

[0037] The Gaussian filtering process refers to an algorithm for smoothing prominent noise points in an image through Gaussian filtering, which is used to reduce noise in the image.

[0038] Specifically, in the embodiment of the present invention, it is first necessary to obtain a fine observation material by cutting, inlaying, grinding, and polishing the titanium alloy material, and then etch it with chemical reagents to obtain the internal structure of the titanium alloy material, and then observe the internal structure using an optical microscope to obtain the microstructure image.

[0039] Specifically, after obtaining the microstructure image, the present invention obtains a denoised tissue image after Gaussian filtering, and then performs feature extraction operation through a convolution pooling process to obtain an image feature set. The present invention classifies the alpha phase set ( phase) and beta phase set ( Finally, the present invention uses edge recognition algorithm to identify each phase and Phase edge recognition is performed, and then the initial two-phase distribution information is obtained. For example, in the region phase and The respective proportions of the phases].

[0040] In detail, in an embodiment of the present invention, obtaining a pre-trained forging annealing recognition model includes: obtaining a historical annealing two-phase change law and a training sample set, obtaining an initialized two-phase distribution recognition network and an initialized annealing recognition network, wherein the initialized annealing recognition network includes an annealing temperature-performance change recognition network and a performance score maximization network; obtaining an annealing temperature law change model, and using the annealing temperature law change model to perform transfer learning on the annealing temperature-performance change recognition network to obtain an updated annealing temperature-performance change recognition network; using the historical annealing two-phase change law, training the updated annealing temperature-performance change recognition network and the performance score maximization network to obtain a trained annealing recognition network; using the training sample set, training the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network; using the trained two-phase distribution recognition network and the annealing recognition network to construct a forging annealing recognition model.

[0041] Among them, the historical annealing two-phase change law refers to the law of annealing temperature and two-phase change obtained from experiments, literature and other materials. For example, the tensile strength and specified non-proportional elongation strength of TC2 titanium alloy bars tend to increase, decrease, increase, and decrease with increasing annealing temperature, and the plasticity tends to decrease and increase with increasing annealing temperature. The impact value, tensile strength and specified non-proportional elongation are the highest at 800°C, and the elongation and cross-sectional shrinkage are the highest at 770°C; the casting will undergo a certain degree of recovery during annealing. When the annealing temperature is 780°C, the structure of the high-oxygen TC4 titanium alloy is more uniform, which is a clustered irregularly distributed lamellar two-phase structure and a small amount of equiaxed structure; as the annealing temperature of TC11 titanium alloy increases, the primary α phase in the alloy decreases or even disappears, while the precipitated needle-like α phase increases, and the alloy changes from a dual-state structure to a fine lamellar structure. Transformation of organizations, etc.

[0042] The training sample set includes the historical initial two-phase distribution of titanium alloy, the historical titanium alloy forging process and the historical two-phase detection of titanium alloy after forging.

[0043] The initialized two-phase distribution identification network is used to analyze the stress generated by forging, and then predict the change of the two-phase distribution based on the stress distribution.

[0044] The initialization annealing recognition network is used to combine a neural network of annealing steps that conform to the change rules based on the existing two-phase distribution and the target two-phase distribution. The initialization annealing recognition network includes an annealing temperature-performance change recognition network and a performance score maximization network. The annealing temperature-performance change recognition network satisfies the above-mentioned overall function, while the performance score maximization network is used to find the most suitable target two-phase distribution.

[0045] The annealing temperature regular change model is a model that has learned the mapping relationship between the annealing temperature and the two-phase changes of the titanium alloy, and can be obtained from the Internet.

[0046] The transfer learning refers to migrating the data parameters in the existing network model to the new network, so that the new network can perform the functions of the existing network model and save training costs.

[0047] Specifically, in an embodiment of the present invention, a historical annealing two-phase change pattern and a training sample set are first obtained, and an initialized two-phase distribution recognition network and an initialized annealing recognition network are obtained. Then, using transfer learning, the annealing temperature-performance change recognition network in the initialized annealing recognition network is transferred using the annealing temperature pattern change model. Since the performance score maximization network in the initialized annealing recognition network is a fixed-parameter network, the training of the annealing temperature-performance change recognition network is completed, and it can be considered that the initialized annealing recognition network has completed preliminary training. Then, training is performed using the historical annealing two-phase change pattern to obtain an annealing recognition network that better identifies the company's titanium alloy materials.

[0048] Furthermore, after the annealing recognition network training is completed, the training sample set can be used to train the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network.

[0049] In detail, in an embodiment of the present invention, the use of the training sample set to train the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network includes: extracting a target sample from the training sample set in sequence, wherein the target sample includes the historical titanium alloy initial two-phase distribution, the historical titanium alloy forging process and the historical titanium alloy forging two-phase detection; using the initialized two-phase distribution recognition network to extract the forging features of the historical titanium alloy forging process to obtain a forging feature set; according to the forging feature set, obtaining a predicted stress distribution, and according to the predicted stress distribution and the historical titanium alloy initial two-phase distribution, predicting a predicted two-phase distribution; calculating the predicted stress distribution. Measure the two-phase distribution and the loss value of the two phases detected after forging the historical titanium alloy, minimize the loss value, obtain the network model parameters when the loss value is minimized, and perform reverse network parameter update on the initialized two-phase distribution identification network according to the network model parameters to obtain an updated two-phase distribution identification network; determine whether the training sample set is a null value; when the training sample set is not a null value, return to the above-mentioned step of extracting a target sample from the training sample set in sequence; when the training sample set is a null value, perform mean calculation based on the network model parameters on the updated two-phase distribution identification network corresponding to each target sample in the training sample set to obtain a trained two-phase distribution identification network.

[0050] The historical initial two-phase distribution of titanium alloy refers to the two-phase distribution of titanium alloy materials produced in the past in an unprocessed state.

[0051] The historical titanium alloy forging process refers to the sequence of forging steps of titanium alloy materials in the past.

[0052] Among them, the two-phase detection after historical titanium alloy forging refers to the two-phase distribution observed through a series of actions such as physical cutting, inlaying, polishing, etc. on titanium alloy forgings manufactured in the past.

[0053] Among them, the forging feature extraction refers to the text feature extraction process about the forging process through a pre-built convolutional layer, and the forging feature set may include applying an upward force to a certain edge area for 3 seconds, causing the titanium alloy to present a 45-degree angle or form a thickness of 3 times, etc.

[0054] Among them, the predicted stress distribution refers to the distribution of the intensity of organizational deformation caused to the titanium alloy by different forging techniques and stress positions. For a normal two-phase distribution, after the influence of the stress distribution, the structure of the two phases in the organization will be destroyed, stretched and displaced.

[0055] Specifically, in the embodiment of the present invention, the two-phase distribution recognition network is initialized by training multiple times and taking the average value.

[0056] During a single training session, the network performs forward computation on the target sample to obtain a predicted two-phase distribution. A cross-entropy loss algorithm is then used to compare the target sample's historical post-forging titanium alloy two-phase detection (equivalent to the true label) with the predicted two-phase distribution (predicted value) to obtain a loss value. The present invention then minimizes this loss value using a gradient descent algorithm to obtain the network model parameters that minimize the loss value. This is then followed by a reverse network parameter update to obtain an updated two-phase distribution recognition network.

[0057] Specifically, in an embodiment of the present invention, the number of target samples in the training sample set is limited. When all target samples in the target sample set are extracted, that is, when the training sample set is null, mean calculation of the network model parameters can be performed to obtain a trained two-phase distribution recognition network. Mean calculation refers to the following: during the training process, the network parameters in the neural nodes are updated as the training progresses. The parameters updated in each training process in a neural node are summed and averaged, thereby obtaining relatively balanced network model parameters in the neural node.

[0058] Furthermore, after the annealing recognition network and the two-phase distribution recognition network are trained, the output layer of the two-phase distribution recognition network can be connected to the input layer of the annealing recognition network to construct a forging annealing recognition model.

[0059] S3. Using the two-phase distribution identification network, perform stress distribution identification operation on the forging process information to obtain a stress distribution matrix, and perform weighted calculation on the initial two-phase distribution information according to the stress distribution matrix to obtain forging two-phase distribution information.

[0060] In detail, in the embodiment of the present invention, the two-phase distribution recognition network is expressed as: , , , where represents the stress distribution matrix, represents the stress identification process, represents the distribution of alloy stress characteristics, Indicates the forging angle of the forging. Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, and Represents the influence coefficient of stress on the two phases.

[0061] Specifically, in the embodiment of the present invention, During the training process, the mapping relationship between forging characteristics and stress distribution is learned, so the distribution of stress characteristics in the forging process information can be and forging bending angle Predictions are made to obtain a stress distribution matrix, which should show the strongest stress distribution in the forged bend area and decrease towards both sides of the bend. Furthermore, the strength of the stress can be determined by the bend radius or angle. For example, a large bend radius results in a small and uniform stress distribution, while a large bend angle indicates a high stress distribution.

[0062] It should be known that different stress distributions have directionality, and different stresses have different Related The influence of the phase is also different, therefore, the present invention is and , and perform weighted calculations on the two-phase distribution respectively, so as to obtain the two-phase distribution information of the forging.

[0063] S4. Using the annealing recognition network, perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and perform a local annealing prediction operation based on two-phase recovery based on the two-phase distribution feature set to obtain a local annealing step and predict the primary recovery two-phase distribution.

[0064] In detail, in an embodiment of the present invention, the annealing recognition network is used to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and a local annealing prediction operation based on two-phase recovery is performed based on the two-phase distribution feature set to obtain a local annealing step and predict the primary recovery two-phase distribution, including: Obtain an objective function in the annealing recognition network, wherein the objective function is expressed as: , where represents the target two-phase distribution, represents the set of annealing schemes, Indicates the permutation and combination of the steps in the annealing scheme set. Representing the existing two-phase distribution; performing a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set; performing a classification operation based on the forging position on the two-phase distribution feature set to obtain the two-phase distribution characteristics of the forging area and the two-phase distribution characteristics of the normal area; according to the objective function, configuring the two-phase distribution characteristics of the forging area as the existing two-phase distribution, configuring the two-phase distribution characteristics of the normal area as the target two-phase distribution, performing a local annealing step identification operation to obtain a local annealing step; using the annealing identification network, according to the local annealing step, obtaining a predicted primary recovery two-phase distribution.

[0065] The objective function needs to be executed twice in the annealing recognition network.

[0066] Specifically, in the embodiments of the present invention, due to the forging process, the folded and bent portions of the forgings can produce two-phase tearing and dispersion, resulting in titanium alloy microstructures in the forged areas having properties far inferior to those in normal areas. Therefore, in the embodiments of the present invention, an annealing recognition network is first used to perform feature extraction operations based on convolution, pooling, and flattening on the two-phase distribution information of the forgings to obtain a two-phase distribution feature set.

[0067] Then, based on the two-phase distribution feature set, the above objective function is executed for the first time, with the two-phase distribution features of the forging area as the existing two-phase distribution and the two-phase distribution features of the normal area as the target two-phase distribution, thereby obtaining a local annealing step sequence.

[0068] Since the annealing recognition network learns the mapping relationship between the annealing steps and the changes in the two-phase distribution through the training process, the predicted primary recovery two-phase distribution can be obtained according to the local annealing steps.

[0069] S5. Perform a global annealing prediction operation on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step.

[0070] It should be noted that it is very demanding to restore the microstructure and properties of the forged folded area to those of the normal area through only local annealing. The difference in microstructure and properties between the forged folded area and the normal area can be further reduced through global annealing.

[0071] Therefore, through global annealing, not only can the microstructure and properties of the forging be developed in the direction of the performance weight distribution, but the microstructure and properties of each region of the forging can also be evenly distributed.

[0072] In detail, in an embodiment of the present invention, a global annealing prediction operation is performed on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step, including: using the annealing identification network, according to the performance weight distribution, a two-phase distribution prediction operation is performed on the initial two-phase distribution information to maximize the performance score to obtain an optimal two-phase distribution; two-phase distribution gap identification is performed on the optimal two-phase distribution and the predicted primary recovery two-phase distribution to obtain a two-phase distribution difference; an annealing step set and an annealing temperature range are obtained, and according to the annealing step set and the annealing temperature range, annealing schemes that meet the two-phase distribution difference are randomly combined to obtain an annealing scheme set; according to a preset annealing cost calculation strategy, the annealing scheme with the lowest annealing cost in the annealing scheme set is screened to obtain an optimal annealing scheme, and a global annealing step corresponding to the optimal annealing scheme is obtained.

[0073] The optimal two-phase distribution refers to the two-phase distribution in the performance organization that best meets the application scenario.

[0074] The two-phase distribution difference refers to the change in the distribution ratio of the two phases per unit area, as well as the change in the relative density of the two phases.

[0075] The annealing step set includes various annealing steps, such as a first-stage annealing at 780°C for 1 hour, a second-stage annealing, and so on, followed by air cooling, furnace cooling, and other steps. A value is extracted from each aspect of the annealing step set and randomly combined to obtain an annealing scheme, such as 780°C for 1.5 hours and AC (air cooling).

[0076] The annealing temperature range refers to the industry standard annealing range for different types of titanium alloys, such as 940-1000 degrees Celsius for TC11 titanium alloy, 730-820 degrees Celsius for TC2 titanium alloy, and 680-830 degrees Celsius for TC4 titanium alloy. The annealing temperatures for different indicators vary, and the union of common indicators can be used to obtain the maximum temperature range.

[0077] Specifically, the present invention executes the objective function in S4 for the second time. This time, the predicted primary recovery two-phase distribution is used as the existing two-phase distribution, and the optimal two-phase distribution composed of the performance weight distribution and the initial two-phase distribution information is used as the target two-phase distribution, thereby obtaining the optimal annealing scheme, and outputting the step sequence in the optimal annealing scheme to obtain the global annealing step.

[0078] S6. According to the local annealing step and the global annealing step, the titanium alloy forging is subjected to a secondary annealing operation to obtain a forging with improved performance.

[0079] In the embodiment of the present invention, a standardized process is adopted, and the local annealing step and the global annealing step are strictly executed through computer control and feedback supervision, thereby obtaining a forging with improved performance.

[0080] In detail, in an embodiment of the present invention, after the titanium alloy forging is subjected to a secondary annealing operation according to the local annealing step and the global annealing step to obtain a performance-enhanced forging, the method further comprises: performing parameter offsets on the local annealing step and the global annealing step according to a preset offset amount to obtain an offset annealing step set; performing an annealing operation on the titanium alloy forging according to the offset annealing step set to obtain an offset annealing forging set, and performing physical testing on the offset annealing forging set to obtain an offset forging test result set; performing physical testing on the performance-enhanced forging The method comprises the following steps: performing a trial operation to obtain a forging test result; calculating the matching degree of the forging test result and the offset forging test result set with the application scenario respectively to obtain a matching score set, and screening the matching score with the largest value in the matching score set to obtain the best matching score; judging whether the best matching score corresponds to the performance-enhanced forging; when the best matching score does not correspond to the performance-enhanced forging, obtaining the offset annealing step corresponding to the best matching score, and training the forging annealing recognition model according to the offset annealing step to obtain an optimized forging annealing recognition model.

[0081] The offset refers to a small-scale modification to the annealing conditions. Different indicators have different offsets, such as a temperature offset of 10 degrees Celsius or a time offset of 10 minutes. Parameter offset refers to adding or subtracting one or more of the offsets to various parameters in the local and global annealing steps.

[0082] Among them, the physical test refers to the process of actually testing the hardness, performance, stability and other properties of forgings through methods such as hydraulic presses, folding machines, corrosive agents, etc.

[0083] Specifically, in an embodiment of the present invention, if the titanium alloy forging is TC2 titanium alloy and the performance weight distribution is such that the requirements for tensile strength and specified non-proportional elongation strength are higher, then the possible global annealing steps are 800°C, 1.5 hours, and air cooling. Based on the offset, a set of offset annealing steps is obtained, such as [800°C, 1.4 hours, air cooling; 790°C, 1.5 hours, air cooling; 790°C, 1.6 hours, air cooling; ...], and thus a set of offset annealed forgings is obtained.

[0084] In this embodiment of the present invention, a set of offset forging test results is obtained through physical testing of tensile strength and specified non-proportional elongation strength: [Forging 1: tensile strength 790, yield strength 655, Forging 2: tensile strength 780, yield strength 650, ...]. Physical testing of the performance-enhanced forging yields a forging test result of [tensile strength 800, yield strength 660], indicating that the performance-enhanced forging still has the best result and passes the test. Conversely, for [Forging x: 800°C, 1.6 hours, air cooling], the test result is [tensile strength 805, yield strength 665]. Based on the offset annealing step of [800°C, 1.6 hours, air cooling], the forging annealing recognition model can be optimized to ensure that the model remains accurate.

[0085] The present invention solves the problem described in the background technology. The present invention first obtains the performance weight distribution of the target according to the application scenario. Through the performance weight distribution, we can know the two-phase ratio that we finally want to adjust the titanium alloy; then obtain the forging process information and initial two-phase distribution information of the titanium alloy forging. Through the forging process information, we can know the deformation of the titanium alloy material, thereby inferring the stress at the titanium alloy forging place, and through machine learning, we can learn the mapping relationship between stress and two-phase distribution changes, so as to predict the two-phase distribution information of the forging after forging; then, the present invention first takes the unforged normal area as the target, and through local annealing, restores the two phases of the forging area as much as possible, and then through global annealing, further reduces the difference between the forging area and the normal area, and can change the two-phase distribution of the titanium alloy during the global annealing process, thereby achieving overall organizational performance improvement. Therefore, the present invention can maximize the organizational performance of the titanium alloy through annealing treatment according to the forging application scenario.

[0086] like Figure 2 1 is a functional module diagram of a system for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to an embodiment of the present invention.

[0087] The system 100 for improving the microstructure and properties of titanium alloy forgings based on annealing temperature described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the system 100 can include a forging performance target acquisition module 101, an information-model acquisition module 102, an annealing step identification module 103, and a step execution module 104. The modules described in the present invention, also referred to as units, are a series of computer program segments that can be executed by an electronic device processor and perform a fixed function, and are stored in the memory of the electronic device.

[0088] The forging performance target acquisition module 101 is used to obtain the application scenario of the pre-built titanium alloy forging, and weight the preset performance indicator set according to the application scenario to obtain the performance weight distribution; the information-model acquisition module 102 is used to obtain the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtain the forging process information of the titanium alloy forging, and obtain the pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network; the annealing step recognition module 103 is used to use the two-phase distribution recognition network to perform stress distribution recognition operation on the forging process information to obtain a stress distribution matrix, and according to the stress The distribution matrix performs weighted calculation on the initial two-phase distribution information to obtain the two-phase distribution information of the forging, and uses the annealing recognition network to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and performs a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution, and performs a global annealing prediction operation on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step; the step execution module 104 is used to perform a secondary annealing operation on the titanium alloy forging according to the local annealing step and the global annealing step to obtain a performance-enhanced forging.

[0089] In detail, the modules in the titanium alloy forging structure performance improvement system 100 based on annealing temperature in the embodiment of the present invention are used in the same manner as above. Figure 1 The same technical means as the method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature described in the previous section can produce the same technical effects, so they will not be repeated here.

[0090] like Figure 3 1 is a schematic diagram of the structure of an electronic device for implementing a method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to an embodiment of the present invention.

[0091] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a method program for improving the microstructure and properties of titanium alloy forgings based on annealing temperature.

[0092] The memory 11 includes at least one type of readable storage medium, including flash memory, a removable hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a plug-in removable hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the electronic device 1. Furthermore, the memory 11 includes both the internal storage unit of the electronic device 1 and an external storage device. The memory 11 can be used not only to store application software installed in the electronic device 1 and various data, such as the code of a program for a method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature, but can also be used to temporarily store data that has been output or is about to be output.

[0093] In some embodiments, the processor 10 may be composed of an integrated circuit, such as a single packaged integrated circuit or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips. The processor 10 is the control core (Control Unit) of the electronic device, connecting the various components of the entire electronic device using various interfaces and circuits. It executes or runs programs or modules stored in the memory 11 (such as a program for improving the microstructure and properties of titanium alloy forgings based on annealing temperature) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0094] The bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to enable communication between the memory 11 and at least one processor 10, etc.

[0095] Figure 3 Only the electronic device with components is shown, and it can be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the electronic device 1 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0096] For example, although not shown, the electronic device 1 may further include a power supply (e.g., a battery) to power various components. Preferably, the power supply may be logically connected to the at least one processor 10 via a power management device, thereby enabling functions such as charge management, discharge management, and power consumption management via the power management device. The power supply may further include any components such as one or more DC or AC power supplies, a recharging device, a power failure detection circuit, a power converter or inverter, and a power status indicator. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which are not further described here.

[0097] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices.

[0098] Optionally, the electronic device 1 may further include a user interface, which may be a display or an input unit (such as a keyboard). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display may also be appropriately referred to as a display screen or a display unit, and is used to display information processed by the electronic device 1 and to display a visual user interface.

[0099] The program of the method for improving the structural performance of titanium alloy forgings based on annealing temperature stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve: obtaining a pre-built application scenario of the titanium alloy forging, and weighting a preset set of performance indicators according to the application scenario to obtain a performance weight distribution; obtaining the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtaining the forging process information of the titanium alloy forging, and obtaining a pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network; using the two-phase distribution recognition network, performing a stress distribution recognition operation on the forging process information , obtain a stress distribution matrix, and perform weighted calculation on the initial two-phase distribution information according to the stress distribution matrix to obtain the two-phase distribution information of the forging; use the annealing recognition network to perform feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and perform a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution; according to the performance weight distribution and the initial two-phase distribution information, perform a global annealing prediction operation on the predicted primary recovery two-phase distribution to obtain a global annealing step; according to the local annealing step and the global annealing step, perform a secondary annealing operation on the titanium alloy forging to obtain a performance-enhanced forging.

[0100] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0101] Furthermore, if the modules / units integrated into the electronic device 1 are implemented as software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. The computer-readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0102] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor of an electronic device, it can achieve the following: obtaining an application scenario of a pre-built titanium alloy forging, and weighting a preset performance indicator set according to the application scenario to obtain a performance weight distribution; obtaining initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtaining forging process information of the titanium alloy forging, and obtaining a pre-trained forging annealing recognition model, wherein the forging annealing recognition model includes a two-phase distribution recognition network and an annealing recognition network; using the two-phase distribution recognition network, performing a stress distribution recognition operation on the forging process information to obtain A stress distribution matrix is ​​obtained, and a weighted calculation is performed on the initial two-phase distribution information according to the stress distribution matrix to obtain the two-phase distribution information of the forging; the annealing identification network is used to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and a local annealing prediction operation based on two-phase recovery is performed according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution; according to the performance weight distribution and the initial two-phase distribution information, a global annealing prediction operation is performed on the predicted primary recovery two-phase distribution to obtain a global annealing step; according to the local annealing step and the global annealing step, a secondary annealing operation is performed on the titanium alloy forging to obtain a performance-enhanced forging.

[0103] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only exemplary, and actual implementations may have other division methods.

[0104] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0105] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0106] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature, characterized in that: The method comprises: obtaining an application scenario of a pre-constructed titanium alloy forging, weighting a preset performance indicator set according to the application scenario, and obtaining a performance weight distribution; obtaining initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, obtaining forging process information of the titanium alloy forging, and obtaining a pre-trained forging annealing recognition model, wherein the forging annealing recognition model comprises a two-phase distribution recognition network and an annealing recognition network; using the two-phase distribution recognition network, performing a stress distribution recognition operation on the forging process information to obtain a stress distribution matrix, and performing a stress distribution recognition operation on the initial two-phase distribution information according to the stress distribution matrix. The information is weightedly calculated to obtain the two-phase distribution information of the forging; the annealing recognition network is used to perform feature extraction on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and a local annealing prediction operation based on two-phase recovery is performed according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution; according to the performance weight distribution and the initial two-phase distribution information, a global annealing prediction operation is performed on the predicted primary recovery two-phase distribution to obtain a global annealing step; according to the local annealing step and the global annealing step, a secondary annealing operation is performed on the titanium alloy forging to obtain a performance-enhanced forging.

2. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 1, characterized in that: The method of weighting a preset performance indicator set according to the application scenario to obtain a performance weight distribution includes: performing text segmentation processing on the application scenario to obtain a text segmentation set, and performing a sequential quantization operation on the text segmentation set to obtain a text quantization set; performing a self-attention enhancement operation on the text quantization set to obtain a text enhancement vector sequence; performing a semantic recognition operation on the text enhancement vector sequence based on a preset performance indicator set to obtain a semantic demand distribution, and performing keyword recognition on the text enhancement vector sequence based on the performance indicator set to obtain a demand word set; performing clustering statistics on the demand word set to obtain a primary demand distribution, and performing weighted calculation on the semantic demand distribution according to the primary demand distribution to obtain a performance weight distribution.

3. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 2, characterized in that: The method of obtaining the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging includes: obtaining a microstructure image of the titanium alloy material of the titanium alloy forging, and performing Gaussian filtering on the microstructure image to obtain a noise-reduced tissue image; performing a feature extraction operation on the noise-reduced tissue image to obtain an image feature set, and performing a two-phase based classification and judgment operation on the image feature set to obtain an alpha phase set and a beta phase set; performing a position frame selection and marking operation on the alpha phase set and the beta phase set to obtain the initial two-phase distribution information.

4. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 3, characterized in that: The method of obtaining a pre-trained forging annealing recognition model includes: obtaining a historical annealing two-phase change law and a training sample set, obtaining an initialized two-phase distribution recognition network and an initialized annealing recognition network, wherein the initialized annealing recognition network includes an annealing temperature-performance change recognition network and a performance score maximization network; obtaining an annealing temperature law change model, and using the annealing temperature law change model to perform transfer learning on the annealing temperature-performance change recognition network to obtain an updated annealing temperature-performance change recognition network; using the historical annealing two-phase change law, training the updated annealing temperature-performance change recognition network and the performance score maximization network to obtain a trained annealing recognition network; using the training sample set, training the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network; and using the trained two-phase distribution recognition network and the annealing recognition network to construct a forging annealing recognition model.

5. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 4, characterized in that: The method of using the training sample set to train the initialized two-phase distribution recognition network to obtain a trained two-phase distribution recognition network includes: extracting a target sample from the training sample set in sequence, wherein the target sample includes the historical initial two-phase distribution of the titanium alloy, the historical titanium alloy forging process, and the two-phase detection after the historical titanium alloy forging; using the initialized two-phase distribution recognition network to extract the forging features of the historical titanium alloy forging process to obtain a forging feature set; obtaining a predicted stress distribution based on the forging feature set, and predicting a predicted two-phase distribution based on the predicted stress distribution and the historical initial two-phase distribution of the titanium alloy; calculating the predicted two-phase distribution and The loss value of the two phases is detected after the historical titanium alloy is forged, and the loss value is minimized to obtain the network model parameters when the loss value is minimized, and the network parameters of the initialized two-phase distribution identification network are reversely updated according to the network model parameters to obtain an updated two-phase distribution identification network; it is determined whether the training sample set is a null value; when the training sample set is not a null value, return to the above-mentioned step of extracting a target sample from the training sample set in sequence; when the training sample set is a null value, the updated two-phase distribution identification network corresponding to each target sample in the training sample set is calculated based on the mean of the network model parameters to obtain a trained two-phase distribution identification network.

6. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 5, characterized in that: The two-phase distribution identification network is expressed as: , , , where represents the stress distribution matrix, represents the stress identification process, represents the stress characteristic distribution of the alloy, Indicates the forging angle of the forging. Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, Indicates the two-phase distribution information of the forging Mutually, Represents the initial two-phase distribution information Mutually, and Represents the influence coefficient of stress on the two phases.

7. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 6, characterized in that: The method of using the annealing recognition network to perform a feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and performing a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and predict a primary recovery two-phase distribution includes: obtaining an objective function in the annealing recognition network, wherein the objective function is expressed as: , where represents the target two-phase distribution, represents the set of annealing schemes, It represents the permutation and combination of the steps in the annealing scheme set. Representing the existing two-phase distribution; performing feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set; performing classification operation based on the forging position on the two-phase distribution feature set to obtain the two-phase distribution characteristics of the forging area and the two-phase distribution characteristics of the normal area; according to the objective function, configuring the two-phase distribution characteristics of the forging area as the existing two-phase distribution, configuring the two-phase distribution characteristics of the normal area as the target two-phase distribution, performing a local annealing step identification operation to obtain a local annealing step; using the annealing identification network, according to the local annealing step, obtaining a predicted primary recovery two-phase distribution.

8. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 7, characterized in that: The method of performing a global annealing prediction operation on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step includes: using the annealing identification network to perform a two-phase distribution prediction operation to maximize the performance score of the initial two-phase distribution information according to the performance weight distribution to obtain an optimal two-phase distribution; identifying the two-phase distribution gap between the optimal two-phase distribution and the predicted primary recovery two-phase distribution to obtain a two-phase distribution difference; obtaining an annealing step set and an annealing temperature range, and randomly combining annealing schemes that meet the two-phase distribution difference according to the annealing step set and the annealing temperature range to obtain an annealing scheme set; and screening the annealing scheme with the lowest annealing cost in the annealing scheme set according to a preset annealing cost calculation strategy to obtain an optimal annealing scheme, and obtaining a global annealing step corresponding to the optimal annealing scheme.

9. The method for improving the microstructure and properties of titanium alloy forgings based on annealing temperature according to claim 8, characterized in that: After performing a secondary annealing operation on the titanium alloy forging according to the local annealing step and the global annealing step to obtain a forging with improved performance, the method further includes: performing parameter offset on the local annealing step and the global annealing step according to a preset offset to obtain an offset annealing step set; performing an annealing operation on the titanium alloy forging according to the offset annealing step set to obtain an offset annealing forging set, and performing physical testing on the offset annealing forging set to obtain an offset forging test result set; performing a physical testing operation on the performance-improved forging to obtain a forging test result; calculating the matching degree between the forging test result and the offset forging test result set and the application scenario respectively to obtain a matching score set, and screening the matching score with the largest value in the matching score set to obtain the best matching score; judging whether the best matching score corresponds to the forging with improved performance; when the best matching score does not correspond to the forging with improved performance, obtaining the offset annealing step corresponding to the best matching score, and training the forging annealing recognition model according to the offset annealing step to obtain an optimized forging annealing recognition model.

10. A titanium alloy forging microstructure and performance improvement system based on annealing temperature, characterized in that: The system comprises: a forging performance target acquisition module, which is used to acquire the application scenario of the pre-built titanium alloy forging, and weight the preset performance indicator set according to the application scenario to obtain the performance weight distribution; an information-model acquisition module, which is used to acquire the initial two-phase distribution information of the titanium alloy material of the titanium alloy forging, acquire the forging process information of the titanium alloy forging, and acquire the pre-trained forging annealing recognition model, wherein the forging annealing recognition model comprises a two-phase distribution recognition network and an annealing recognition network; an annealing step recognition module, which is used to use the two-phase distribution recognition network to perform stress distribution recognition operation on the forging process information to obtain the stress distribution matrix, and obtain the stress distribution matrix according to the stress distribution matrix. The distribution matrix performs weighted calculation on the initial two-phase distribution information to obtain the two-phase distribution information of the forging, and uses the annealing recognition network to perform feature extraction operation on the two-phase distribution information of the forging to obtain a two-phase distribution feature set, and performs a local annealing prediction operation based on two-phase recovery according to the two-phase distribution feature set to obtain a local annealing step and a predicted primary recovery two-phase distribution, and performs a global annealing prediction operation on the predicted primary recovery two-phase distribution according to the performance weight distribution and the initial two-phase distribution information to obtain a global annealing step; a step execution module is used to perform a secondary annealing operation on the titanium alloy forging according to the local annealing step and the global annealing step to obtain a performance-enhanced forging.

Images