A method, system and electronic device for measuring primary alpha phase grain size of a titanium alloy
By constructing a high-dimensional multi-parameter ultrasonic evaluation model and optimizing the parameters using a differential evolution algorithm, the problems of low accuracy and efficiency in measuring the primary α phase grain size of titanium alloys were solved, achieving higher accuracy and robustness in measurement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANCHANG HANGKONG UNIVERSITY
- Filing Date
- 2023-05-12
- Publication Date
- 2026-05-12
AI Technical Summary
Existing ultrasonic evaluation models for primary α phase grain size in titanium alloys suffer from insufficient evaluation accuracy and low solution efficiency, making it difficult to effectively explain the coupling between parameters and the large computational load.
A high-dimensional model is used to represent a multi-parameter ultrasonic evaluation model. The model is trained using training samples, and the undetermined coefficients are solved by differential evolution algorithm. An optimization problem is established with the goal of minimizing the average error of the primary α phase grain size. The model parameters are optimized by combining differential grouping and co-evolution algorithm.
It improves the accuracy and robustness of primary α phase grain size measurement, reduces measurement errors, and has good anti-interference and solution efficiency.
Smart Images
Figure CN116448022B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of primary α phase grain size measurement in titanium alloys, and in particular to a method, system, and electronic device for measuring the primary α phase grain size in titanium alloys. Background Technology
[0002] Titanium alloys possess excellent corrosion and oxidation resistance, making them widely used in the manufacture of aero-engine casings and turbines. During the manufacturing process, the microstructure of titanium alloys is easily affected by temperature, inevitably leading to wear and deformation. The primary α-phase grain size is a crucial parameter characterizing the microstructure of titanium alloys. To ensure the reliability of alloy quality, the detection of the primary α-phase grain size under different temperatures and deformations is extremely important. Ultrasonic testing, with its excellent penetration capability, can perform localized and quantitative detection, offering high sensitivity and resolution, and has been widely used in the non-destructive evaluation of primary α-phase grain size.
[0003] The core of ultrasonic evaluation technology for titanium alloys is to determine a quantitative relationship model between the size of primary α-phase grains and ultrasonic characteristic parameters. Essentially, this involves fitting a mathematical function model of ultrasonic parameters and the size of primary α-phase grains. When ultrasound propagates in titanium alloys, it is affected by the complex internal microstructure. If a single ultrasonic characteristic parameter, such as ultrasonic velocity, attenuation coefficient, nonlinear coefficient, or first-order backwave amplitude, is used to model and evaluate the size of primary α-phase grains, the model's evaluation accuracy will be insufficient due to the limited amount of ultrasonic information. If multiple ultrasonic characteristic parameters are used simultaneously to evaluate the size of primary α-phase grains, multi-parameter model construction becomes difficult. Low-complexity models cannot effectively explain the coupling between parameters, while high-complexity models have high computational cost and low solution efficiency. The dimension of undetermined coefficients in these models is often hundreds or even thousands, significantly increasing the difficulty of determining the undetermined coefficients within the evaluation model. Solving optimization problems with high-dimensional and ultra-high-dimensional decision variables presents a new challenge to the quantitative evaluation of the size of primary α-phase grains and ultrasonic characteristic parameters in titanium alloys.
[0004] Therefore, current ultrasonic evaluation models and methods for solving model parameters for primary α phase grain size suffer from insufficient model evaluation accuracy and low solution efficiency. Summary of the Invention
[0005] The purpose of this invention is to provide a method, system, and electronic device for measuring the size of primary α-phase grains in titanium alloys, so as to improve the accuracy of primary α-phase grain size measurement.
[0006] To achieve the above objectives, the present invention provides the following solution:
[0007] A method for measuring the primary α-phase grain size of titanium alloys, comprising:
[0008] Obtain the ultrasonic characteristic parameters of titanium alloys;
[0009] The ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters.
[0010] Based on the normalized ultrasonic characteristic parameters, the primary α-phase grain size of the titanium alloy is obtained using the primary α-phase grain size measurement model.
[0011] The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys obtained by scanning with water immersion probes at different frequencies and the corresponding actual values of primary α-phase grain size. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
[0012] Optionally, the ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters, specifically including:
[0013] Using formula The ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters; wherein, These are ultrasound characteristic parameters. Let x be the normalized ultrasound feature parameters, max(Z) be the maximum value of the ultrasound feature parameters input in the current dimension, and min(Z) be the minimum value of the ultrasound feature parameters input in the current dimension. L x is the lower bound of the input ultrasound feature parameters. U The upper bound of the input ultrasound characteristic parameters.
[0014] Optionally, the construction of the initial measurement model for the primary α-phase grain size specifically includes:
[0015] Constructing an initial measurement model for primary α-phase grain size using a high-dimensional model Among them, f i (x i ) represents the first-order function expression of a high-dimensional model; x i f is the i-th ultrasound characteristic parameter; ij (x i ,x j Let x be the second-order function expression of the high-dimensional model; j Let j be the j-th ultrasound feature parameter; n is the number of ultrasound feature parameters.
[0016] Optionally, the process of constructing the primary α-phase grain size measurement model specifically includes:
[0017] The undetermined coefficients of the initial measurement model of the primary α phase grain size are solved using the differential evolution algorithm, and the second measurement model of the primary α phase grain size under the current iteration number is obtained.
[0018] The ultrasonic characteristic parameters of the titanium alloy obtained by scanning with a water immersion probe at each frequency are input into the second measurement model of the primary α phase grain size under the current iteration number to obtain multiple predicted values of the primary α phase grain size under the current iteration number.
[0019] Calculate the error between the predicted value of the primary α phase grain size and the actual value of the primary α phase grain size at each current iteration number, and calculate the average of multiple errors to obtain the average error at the current iteration number;
[0020] Determine if the maximum number of iterations has been reached;
[0021] If not, the second measurement model of the primary α phase grain size is used as the initial measurement model of the primary α phase grain size, and the process returns to the step of "using the differential evolution algorithm to solve the undetermined coefficients of the initial measurement model of the primary α phase grain size to obtain the second measurement model of the primary α phase grain size under the current iteration number" to proceed to the next iteration.
[0022] If so, the second measurement model of the primary α phase grain size under the number of iterations corresponding to the minimum average error shall be used as the measurement model of the primary α phase grain size.
[0023] A system for measuring the grain size of primary α phase in titanium alloys, comprising:
[0024] The data acquisition module is used to acquire the ultrasonic characteristic parameters of the titanium alloy;
[0025] The normalization processing module is used to normalize the ultrasound feature parameters to obtain normalized ultrasound feature parameters.
[0026] The measurement module is used to obtain the primary α phase grain size of the titanium alloy based on the normalized ultrasonic characteristic parameters and using the primary α phase grain size measurement model.
[0027] The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys obtained by scanning with water immersion probes at different frequencies and the corresponding actual values of primary α-phase grain size. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
[0028] An electronic device includes: a memory and a processor, wherein the memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the above-described method for measuring the primary α-phase grain size of titanium alloys.
[0029] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for measuring the primary α-phase grain size of titanium alloys.
[0030] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0031] This invention provides a method, system, and electronic device for measuring the primary α-phase grain size of titanium alloys. The method first obtains the ultrasonic characteristic parameters of the titanium alloy; then, it normalizes these parameters to obtain normalized ultrasonic characteristic parameters; based on these normalized ultrasonic characteristic parameters, it uses a primary α-phase grain size measurement model to obtain the primary α-phase grain size of the titanium alloy. The primary α-phase grain size measurement model is trained using training samples, with the optimization objective being to minimize the average error between the actual and predicted primary α-phase grain sizes. The undetermined coefficients of the initial primary α-phase grain size measurement model are obtained using a differential evolution algorithm. This initial primary α-phase grain size measurement model is constructed using a high-dimensional model. The method of this invention yields a smaller error in the measured primary α-phase grain size, exhibits good robustness and strong anti-interference capabilities, and solves the problems of low model measurement accuracy and low solution efficiency. Attached Figure Description
[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0033] Figure 1 Flowchart of the method for measuring the primary α phase grain size of titanium alloys provided by the present invention;
[0034] Figure 2 This is a flowchart illustrating the practical application of the method for measuring the primary α phase grain size of titanium alloys according to the present invention.
[0035] Figure 3 Metallographic images of TC25 titanium alloy at different forging temperatures and deformation amounts;
[0036] Figure 4The training sample fitting curve of the method for measuring the primary α phase grain size of titanium alloys;
[0037] Figure 5 A graph showing the relationship between the measurement method of primary α phase grain size in titanium alloys and the average value of primary α phase grain size;
[0038] Figure 6 This is a statistical chart showing the proportion of errors in the test samples. Detailed Implementation
[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0040] The purpose of this invention is to provide a method, system, and electronic device for measuring the size of primary α-phase grains in titanium alloys, so as to improve the accuracy of primary α-phase grain size measurement.
[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0042] Example 1
[0043] This embodiment discloses a method for measuring the primary α phase grain size of titanium alloys. It proposes a multi-parameter ultrasonic evaluation model (initial measurement model of primary α phase grain size) expressed by a high-dimensional model. This method can solve the problem of multi-parameter model construction and the correlation between reasonable response parameters. With the rise of intelligent computing, the model parameters are solved by establishing the target direction with the minimum average error. This provides a new technical means for evaluating the primary α phase grain size using a multi-parameter model.
[0044] In this embodiment, a multi-parameter ultrasonic evaluation model is constructed using eight ultrasonic characteristic parameters, including sound velocity, attenuation coefficient, first bottom wave amplitude, second bottom wave amplitude, and nonlinear coefficient, with the goal of improving the accuracy of primary α-phase grain size evaluation. All experimental data are normalized, all training samples are input into the model, and the undetermined coefficients of the model are transformed into decision variables. A single-objective optimization problem is established to minimize the average error of primary α-phase grain size in the training samples. The decision variables are solved using a differential grouping co-evolutionary algorithm (differential evolution algorithm) to determine the final evaluation model (primary α-phase grain size measurement model).
[0045] like Figure 1 As shown, the method for measuring the primary α-phase grain size of titanium alloys provided by the present invention includes:
[0046] Step 101: Obtain the ultrasonic characteristic parameters of the titanium alloy.
[0047] Step 102: Normalize the ultrasound characteristic parameters to obtain normalized ultrasound characteristic parameters.
[0048] As an optional embodiment, step 102 specifically includes:
[0049] The ultrasound characteristic parameters are normalized using formula (1) to obtain the normalized ultrasound characteristic parameters.
[0050]
[0051] in, These are ultrasound characteristic parameters. Let x be the normalized ultrasound feature parameters, max(Z) be the maximum value of the ultrasound feature parameters input in the current dimension, and min(Z) be the minimum value of the ultrasound feature parameters input in the current dimension. L x is the lower bound of the input ultrasound feature parameters. U The upper bound of the input ultrasound characteristic parameters.
[0052] Step 103: Based on the normalized ultrasonic characteristic parameters, the primary α-phase grain size of the titanium alloy is obtained using the primary α-phase grain size measurement model.
[0053] The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys obtained by scanning with water immersion probes at different frequencies and the corresponding actual values of primary α-phase grain size. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
[0054] As an optional embodiment, the process of constructing the primary α-phase grain size measurement model specifically includes:
[0055] The undetermined coefficients of the initial measurement model of the primary α phase grain size are solved using the differential evolution algorithm, and the second measurement model of the primary α phase grain size under the current iteration number is obtained.
[0056] The ultrasonic characteristic parameters of the titanium alloy obtained by scanning with a water immersion probe at each frequency are input into the second measurement model of the primary α phase grain size under the current iteration number to obtain multiple predicted values of the primary α phase grain size under the current iteration number.
[0057] Calculate the error between the predicted value of the primary α phase grain size and the actual value of the primary α phase grain size for each current iteration number, and calculate the average of multiple errors to obtain the average error for the current iteration number.
[0058] Determine if the maximum number of iterations has been reached.
[0059] If not, the second measurement model of the primary α phase grain size is used as the initial measurement model of the primary α phase grain size, and the process returns to the step of "using the differential evolution algorithm to solve the undetermined coefficients of the initial measurement model of the primary α phase grain size to obtain the second measurement model of the primary α phase grain size under the current iteration number" to proceed to the next iteration.
[0060] If so, the second measurement model of the primary α phase grain size under the number of iterations corresponding to the minimum average error shall be used as the measurement model of the primary α phase grain size.
[0061] This invention uses TC25 titanium alloy as a specific embodiment to describe in detail the method for measuring the primary α phase grain size of titanium alloy.
[0062] like Figure 2 As shown, the steps of the method for measuring the primary α phase grain size of titanium alloys according to the present invention in practical applications are as follows:
[0063] S1. The grain boundaries of the TC25 titanium alloy sample were extracted through forging to obtain the size of the primary α phase grains. Then, an ultrasonic signal was generated using a 5077PR pulse signal generator to excite pulses. Under different frequencies of water immersion probes, corresponding ultrasonic waves were emitted using water as the propagation medium. The ultrasonic scanning signals were acquired using a PicoScope3027B acquisition card, and ultrasonic characteristic parameters were extracted. The ultrasonic characteristic parameters include ultrasonic velocity, velocity variance, ultrasonic attenuation coefficient, attenuation coefficient variance, ultrasonic nonlinearity coefficient, first-order backwave shift, second-order backwave shift, and fundamental wave amplitude.
[0064] S2. Normalize the ultrasound characteristic parameters, construct a multi-parameter ultrasound evaluation model using a high-dimensional model, and treat the undetermined coefficients in the evaluation model as decision variables to determine the magnitude of the decision variables. This includes the following steps:
[0065] S21. To maintain dimensional consistency of the ultrasound characteristic parameters, the ultrasound characteristic parameters are normalized to 0.001 to 0.099, and 0.001 is considered as the lower bound of the input ultrasound characteristic parameters, denoted as x. L Let 0.999 be considered the upper bound, denoted as x. U .
[0066] S22. Construct a high-dimensional multi-ultrasound parameter evaluation model. The high-dimensional model expression construction method is adopted, and its specific expression is Equation (2):
[0067]
[0068] In the formula, n is the dimension of the input variable (the number of ultrasound feature parameters), f i (x i The function expression is as follows (3), f ij (x i ,x j The function expression is as follows (4).
[0069]
[0070] In the formula, f i (x i The undetermined coefficients of the function. It is 0.005. It is 0.001. x is 0.999. i These are the ultrasound feature parameters to be input.
[0071]
[0072] In the formula, and f ij (x i ,x j The undetermined coefficients of the function. and It is 0.005. and It is 0.999. and x is 0.001. i With x j These are the different ultrasound feature parameters to be input.
[0073] S23. Group the sample data into training samples and test samples according to the process number and an 8:2 ratio. Determine the number of input ultrasonic characteristic parameter variables, calculate the dimension of the undetermined coefficients of the objective function multi-ultrasonic parameter evaluation model, i.e., the number of decision variables. The sample coefficient T is calculated as follows:
[0074]
[0075] In the formula, C i and C ij To evaluate the undetermined coefficients of the model.
[0076] The number of ultrasound characteristic parameters involved in the evaluation was 8, and the calculated dimension of the decision variables was 276.
[0077] S3. Input all training samples into the constructed evaluation model and establish a single-objective optimization problem with the goal of minimizing the average error of the primary α phase grain size. The coefficients within the model are solved using a differential grouping co-evolutionary algorithm. The decision variable coefficients of the ultrasonic evaluation model are solved using an optimization strategy that aims to minimize the average error of the fitted value of the primary α phase grain size.
[0078] S3 specifically includes:
[0079] S31. Substitute all training samples one by one into the multi-parameter ultrasound evaluation model and sum them to form the objective function. Take the absolute value of the objective function. The objective function is calculated as follows:
[0080]
[0081] In the formula, Equation (2) above, where N1 is the number of training samples.
[0082] S32. After the ultrasound characteristic parameter values are input into the objective function, the undetermined coefficients of the model are converted into decision variables. An optimization objective function is established, and the errors between the fitted values and the true values of all training samples are accumulated and added together. The average value of this sum is then calculated, which is the mean absolute error. The calculation formula is as follows.
[0083]
[0084] S33. Initialize the population size of the objective function solution vectors. Based on the fitness value of each individual in the population, randomly and uniformly generate solution vectors. Each solution vector has T dimensions, and set the range of values for each dimension. For example, if one solution is X... i =(x i,1 ,x i,2 ,x i,3 ,…x i,T Set the maximum number of evolutionary iterations gen, the mutation factor F, and the crossover factor P during evolution.
[0085] S34. For large-scale optimization problems of the objective function, perform differential grouping to break them down into several smaller subproblems. The grouping method uses the first decision variable as the baseline and progressively checks for interactions with other decision variables. If the algorithm detects an interaction, the current decision variable is excluded from all variables and added to the first subcomponent. This process is repeated until all decision variables and the baseline decision variable are excluded, forming the first subcomponent. If no interaction is detected, the variable is considered separable. This process is repeated until the baseline variable is the last decision variable. The interaction detection formula is:
[0086]
[0087] In the formula: f is a separable function, x p x q As decision variables, For decision variable x of f p The positive difference δ, where a is the decision variable x. p The random values that can be obtained; b1 and b2 are the decision variables x. q Two different random values can be obtained.
[0088] The interaction detection method is to fix the value of the first decision variable (here, the first decision variable is represented as x). p (The value is a), and the values of other decision variables (here, the other decision variables are x) are changed. q Let q = 1, 2, 3, ..., T, and p and q are not equal. Let x... q The value changes from b1 to b2), and we check if their delta values are equal. If they are not equal, there is an interaction effect, and x is adjusted accordingly. p Place the first subcomponent a (i.e., x) p With x q The changes are independent; if not independent, then there is an interaction.
[0089] S35. Randomly select 3 solution vectors from the initialized population, denoted as X. i ,X j ,X k , will X i Consider it as the objective solution vector, and the solution vector X j and X k Perform a mutation operation to generate a mutated solution.
[0090] S36. Cross the mutated solution vector with the solution vector according to the crossover factor probability to select a new target solution. The crossover calculation formula is as follows:
[0091]
[0092] In the formula, For the variant solution, X i X represents the current solution, and X represents the target solution selected after the crossover.
[0093] S37. Collaborate on several subproblems to generate a complete solution to the objective function, calculate the fitness value of the newly generated objective solution vector, compare and select until the maximum population size is reached, repeat the evolution operation gen times until the average error is minimized.
[0094] S38. When the objective function value is calculated to the minimum value, the model decision variables are solved, the current complete solution vector is recorded, the undetermined coefficients of the multi-ultrasound parameter evaluation model are determined, and the final model is obtained.
[0095] In Example 1, a total of 160 experimental samples were extracted using the method described in S1. The microstructure of the samples obtained by metallographic microscopy is shown below. Figure 3 As shown in the figure. The experimental sample data was divided into 128 training samples and 32 test samples. All training samples were input into a multi-parameter ultrasonic evaluation model constructed using the S2 method high-dimensional model expression. The optimization objective was to minimize the average error of the fitted values of the primary α phase grain size. The problem was transformed into the problem of determining the undetermined coefficients of the model. The S3 difference grouping co-evolutionary algorithm was used to solve for the undetermined coefficients, and the final evaluation model was calculated. Finally, the model method was verified and analyzed using the S4 method on the test samples. The comparison between the fitted values of the primary α phase grain size of the training samples and the true values of the test samples is shown in the figure. Figure 4 and Figure 5 As shown in Table 1, the evaluation index values of the test sample results are shown in Table 2, and the experimental results are shown in Table 2.
[0096] Table 1. Statistical Table of Evaluation Indicator Values
[0097]
[0098] Table 232 Statistical Table of Experimental Results for Test Samples
[0099]
[0100] The data in Table 1 shows that the method of establishing a single-objective optimization direction and using a differential grouping co-evolutionary algorithm to solve for the undetermined coefficients within a multi-parameter ultrasonic evaluation model has an average absolute error of 0.84, a root mean square error of 1.15, and a coefficient of determination of 0.80 for the test sample evaluation results. The results indicate that, compared to the single-parameter evaluation method, this invention adds ultrasonic characteristic parameters for modeling, fully utilizing ultrasonic acoustic information. Although the undetermined coefficients of the model constitute a large-scale optimization problem, the use of the differential evolutionary algorithm results in a smaller error and higher accuracy in evaluating the primary α-phase grain size. The method is effective, feasible, and has better robustness. The percentage error of the test samples is statistically segmented as follows: Figure 6 As shown, it can be seen that among the 32 test samples, 16 groups have an error of 0-5%, 12 groups have an error of 5-10%, and the remaining 4 groups have an error of 4%. This proves that expanding the input parameters and using the differential grouping co-evolutionary algorithm to solve the model coefficients is effective and feasible without losing acoustic information.
[0101] This invention employs a method for measuring the primary α-phase grain size in titanium alloys. It utilizes a high-dimensional model to establish an initial measurement model for the primary α-phase grain size, transforming the problem into solving for undetermined coefficients within the model. The evaluation optimization objective is to minimize the average absolute error. Finally, a differential evolution algorithm is used to solve the model, yielding the final ultrasonic evaluation model. This method addresses the problems of insufficient utilization of acoustic information in single-parameter methods and the difficulty in solving complex undetermined coefficients in multi-parameter models. It is an effective model-solving method for evaluating the primary α-phase grain size in titanium alloys.
[0102] Example 2
[0103] To implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, a primary α-phase grain size measurement system for titanium alloys is provided below, comprising:
[0104] The data acquisition module is used to acquire the ultrasonic characteristic parameters of titanium alloys.
[0105] The normalization module is used to normalize the ultrasound feature parameters to obtain normalized ultrasound feature parameters.
[0106] The measurement module is used to obtain the primary α phase grain size of the titanium alloy based on the normalized ultrasonic characteristic parameters and using the primary α phase grain size measurement model.
[0107] The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys obtained by scanning with water immersion probes at different frequencies and the corresponding actual values of primary α-phase grain size. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
[0108] Example 3
[0109] This embodiment provides an electronic device, including: a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the method for measuring the primary α phase grain size of titanium alloys according to Embodiment 1.
[0110] Example 4
[0111] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for measuring the primary α phase grain size of titanium alloys according to Embodiment 1.
[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.
[0113] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A method for measuring the size of primary α-phase grains in titanium alloys, characterized in that, include: Obtain the ultrasonic characteristic parameters of titanium alloys; The ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters. Based on the normalized ultrasonic characteristic parameters, the primary α-phase grain size of the titanium alloy is obtained using the primary α-phase grain size measurement model. The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys and actual primary α-phase grain size values obtained by scanning with water immersion probes at different frequencies. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
2. The method for measuring the primary α phase grain size of titanium alloys according to claim 1, characterized in that, The ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters, specifically including: Using formula The ultrasound characteristic parameters are normalized to obtain normalized ultrasound characteristic parameters; wherein, These are ultrasound characteristic parameters. These are the normalized ultrasound characteristic parameters. This represents the maximum value of the ultrasound feature parameters input in the current dimension. This represents the minimum value of the ultrasound feature parameters input in the current dimension. This represents the lower bound of the input ultrasound feature parameters. The upper bound of the input ultrasound characteristic parameters.
3. The method for measuring the primary α phase grain size of titanium alloys according to claim 1, characterized in that, The construction of the initial measurement model for the primary α-phase grain size specifically includes: Constructing an initial measurement model for primary α-phase grain size using a high-dimensional model ;in, This is the first-order function expression for a high-dimensional model; Let i be the i-th ultrasound feature parameter; This is a second-order function expression for a high-dimensional model; Let j be the j-th ultrasound feature parameter; n is the number of ultrasound feature parameters.
4. The method for measuring the primary α phase grain size of titanium alloys according to claim 1, characterized in that, The construction process of the primary α-phase grain size measurement model specifically includes: The undetermined coefficients of the initial measurement model of the primary α phase grain size are solved using the differential evolution algorithm, and the second measurement model of the primary α phase grain size under the current iteration number is obtained. The ultrasonic characteristic parameters of the titanium alloy obtained by scanning with a water immersion probe at each frequency are input into the second measurement model of the primary α phase grain size under the current iteration number to obtain multiple predicted values of the primary α phase grain size under the current iteration number. Calculate the error between the predicted value of the primary α phase grain size and the actual value of the primary α phase grain size at each current iteration number, and calculate the average of multiple errors to obtain the average error at the current iteration number; Determine if the maximum number of iterations has been reached; If not, the second measurement model of the primary α phase grain size is used as the initial measurement model of the primary α phase grain size, and the process returns to the step of "using the differential evolution algorithm to solve the undetermined coefficients of the initial measurement model of the primary α phase grain size to obtain the second measurement model of the primary α phase grain size under the current iteration number" to proceed to the next iteration. If so, the second measurement model of the primary α phase grain size under the number of iterations corresponding to the minimum average error shall be used as the measurement model of the primary α phase grain size.
5. A system for measuring the grain size of primary α phase in titanium alloys, characterized in that, include: The data acquisition module is used to acquire the ultrasonic characteristic parameters of the titanium alloy; The normalization processing module is used to normalize the ultrasound feature parameters to obtain normalized ultrasound feature parameters. The measurement module is used to obtain the primary α phase grain size of the titanium alloy based on the normalized ultrasonic characteristic parameters and using the primary α phase grain size measurement model. The primary α-phase grain size measurement model is trained using training samples. The optimization objective is to minimize the average error between the actual and predicted primary α-phase grain size values. The undetermined coefficients of the primary α-phase grain size measurement model are obtained by solving the differential evolution algorithm. The training samples include ultrasonic characteristic parameters of titanium alloys and actual primary α-phase grain size values obtained by scanning with water immersion probes at different frequencies. The primary α-phase grain size measurement model is constructed using a high-dimensional model.
6. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to cause the electronic device to perform the method for measuring the primary α phase grain size of titanium alloys according to any one of claims 1-4.
7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method for measuring the primary α phase grain size of titanium alloys according to any one of claims 1-4.