Method and system for optimizing shape of high-temperature alloy disc type pre-forged piece

Abstract

The invention belongs to the technical field of material science, and discloses a shape optimization method for a high-temperature alloy disc type pre-forged piece. According to the method, finite element numerical simulation, a support vector machine and a genetic algorithm are combined, a GA-SVR prediction model between an input variable and a target function is established, the optimal value of the target function is predicted, and the optimal geometric parameter combination of the pre-forged piece is obtained. Firstly, the shape of a pre-forged piece is preliminarily designed through experience, then a finite element simulation result is extracted to establish a data set, parameters of a support vector machine model serve as a target function of a genetic algorithm by means of the data set, the parameters of the support vector machine model are optimized by means of the genetic algorithm, and an optimal prediction model is obtained. And for the size parameter combination of the pre-forged piece when the best deformation uniformity evaluation index is obtained, the deformation uniformity of finish forging is effectively improved through global optimization of the genetic algorithm, so that the final performance of the forged piece is improved.

Description

Technical Field

[0001] The present invention belongs to the field of material science and technology, and in particular relates to a method and system for optimizing the shape of a high-temperature alloy disc pre-forging. Background Art

[0002] As the key hot end components of aircraft engines, high-temperature alloy disc forgings need to work under extremely high temperature, high pressure, high speed and complex load environment, and the requirements for their performance and dimensional accuracy are extremely strict. At present, the die forging process of turbine discs faces problems such as filling difficulties, uneven deformation and coarse grains, which makes it difficult to meet the quality requirements of forgings. Therefore, it is necessary to optimize the organizational density and uniformity of forgings to improve their reliability and stability during service. Due to the complex structure of disc forgings, the forging process usually requires multiple processes such as roughing, pre-forging and final forging. The shape of the pre-forging has an important influence on the flow and distribution of the material in the die during the final forging stage, thereby determining the dimensional accuracy, deformation uniformity and microstructural state of the formed forgings.

[0003] Existing pre-forging design methods are usually based on experience or analog design of existing die forgings. Physical experiments are used to verify the feasibility of pre-forging schemes. Although they are practical, they are inefficient. With the application of finite element numerical simulation technology, the cost of scheme verification, structural optimization and process parameter calibration in forging development has been greatly reduced, and simulation methods have been widely used in the optimization process. However, for highly complex pre-forging geometric design, it is obviously difficult to meet the multi-objective optimization requirements by relying solely on empirical design and finite element simulation.

[0004] In current technology, in order to optimize the performance of forgings and effectively control grain size and deformation uniformity, global optimization methods such as response surface method, neural network and sensitivity analysis are generally used. Through these methods, the relationship between design variables and objective functions is established in quantitative analysis, making multi-objective optimization design more efficient and in line with process requirements. However, the construction of response surface models in existing methods is relatively cumbersome. For example, PB experimental design and rotation combination design both require manual statistics and variance calculation, which is time-consuming and labor-intensive. In addition, in order to obtain the experimental response value, each set of parameters needs to be remodeled and imported into the simulation, which is a cumbersome process; and although the second-order response surface model is suitable for simulating curve relationships, it has poor robustness when dealing with highly nonlinear problems and is easily disturbed by outliers. The insufficient number of experimental schemes makes it difficult for the final model to accurately capture the complex interactions and nonlinear characteristics between multiple variables, affecting the optimization efficiency.

[0005] Therefore, the problems and shortcomings of the existing technology are that the traditional forging process cannot meet the requirements of high precision and high performance of turbine disks, and the existing response surface method lacks an accurate design method for optimizing the design of pre-forging dies. It is urgent to develop a new method based on efficient pre-forging shape optimization, combining parametric geometric design, improved experimental design and optimization model to improve forging performance, shorten production cycle, and reduce production costs, so as to meet the demanding service requirements of high-temperature alloy disk forgings in aerospace engines. Summary of the invention

[0006] In view of the problems existing in the prior art, the present invention provides a method for optimizing the shape of a high-temperature alloy disc pre-forging.

[0007] The present invention is implemented in this way: a method for optimizing the shape of a high-temperature alloy disc pre-forging comprises:

[0008] Step 1: Simulate the disc-like parts without pre-forging process. According to the overall equivalent strain of the disc-like parts after final forging, use the reverse simulation technology to quickly design the initial geometric shape of the pre-forging parts, and design the shape of the pre-forging die according to the shape of the pre-forging parts;

[0009] Step 2: By combining the experimental design method with the finite element method, a screening experimental design is performed to determine the characteristic geometric parameters of the pre-forging in step 1 as the key optimization parameters, and determine the value range of the optimization parameters;

[0010] Step 3: Use the Latin hypercube sampling method to generate approximate random samples within the range of multivariate parameter values, set the number of samples drawn by the Latin hypercube sampling, and use the finite element simulation method to automatically batch simulate all random samples;

[0011] Step 4: Establish an objective function that describes the forming quality, and extract the required data from all sample simulation results in step 3 as a data set for establishing a prediction model;

[0012] Step 5: A GA-SVR prediction model between the input variables and the objective function is established through the data set in step 4, the optimal value of the objective function is predicted, and the best combination of pre-forging geometric parameters is obtained.

[0013] Furthermore, the simulation of the disc-like parts without pre-forging process is performed, and according to the overall effect deformation of the disc-like parts after final forging, the initial geometric shape of the pre-forging parts is quickly designed by using the reverse simulation technology, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts:

[0014] The final forging finite element simulation is directly performed on the billet. The forging effect strain results obtained by the finite element simulation are used to reversely trace the positions of the small deformation zone and the large deformation zone in the original billet, and the preliminarily designed shape of the pre-forging is designed. This method can eliminate the small deformation zone and the large deformation zone, and can significantly improve the deformation uniformity of the forging. The specific design method is to first divide the volume phase into N rectangular areas along the radial direction according to the deformation after the final forging. If the shape of the final forging is more complex, the number of partitions for the final forging is also greater. Then, according to the partitions for the final forging, the billet is divided into N rectangular areas of the same volume; a script is written to extract the strain value ∈i of each grid after the final forging, and the average effect strain value of the forging after the final forging is calculated. Then, the deformation of the large deformation area and the small deformation area of ​​the final forging is analyzed, and the small deformation area and the large deformation area smaller than the forging are marked, and the reverse tracking method is used to track the positions of the small deformation area and the large deformation area in the original billet. The points that need to be tracked after the final forging are selected by point tracking in the post-processing process of the finite element simulation software, and the positions of the selected points in the billet are tracked by reverse simulation. Then, the material of the billet and its corresponding small deformation area is reduced, and the reduced volume is added to the large deformation area along the radial direction. Finally, the sharp part of each different rectangular area is smoothed to obtain the initial pre-forging shape;

[0015] Mean effect size

[0016] ∈i is the effect strain of the ith grid, n is the total number of grids;

[0017]

[0018] α and β are parameters for selecting the overall deformation, and their values ​​range from 0.1 to 0.5.

[0019] Furthermore, the experimental design method is combined with the finite element method to perform a screening experimental design to determine the characteristic geometric parameters of the pre-forging in step 1 as key optimization parameters, and determine the value range of the optimization parameters;

[0020] The geometric dimensions that mainly affect the effect deformation of the forging during the pre-forging process are selected as the key geometric dimensions. The selection method is to extract all the geometric parameters of the pre-forging die preliminarily designed in step 1, and then set all the parameters as design variables of the finite element simulation. The experimental design method is combined with the finite element method, and a screening experimental design is performed to evaluate the influence of the geometric dimensions of the die on the effect deformation of the forging. In the post-processing of the finite element simulation, a relationship diagram of the influence of all geometric parameters on the effect deformation is obtained. The Spearman correlation coefficient (rs) is used to evaluate the influence of each geometric parameter variable on the effect deformation. When the Spearman correlation coefficient is greater than q, the parameter is selected as the characteristic geometric parameter. q is the threshold value of the geometric parameter having a greater influence on the effect deformation, ranging from 0.1 to 0.3. The geometric dimensions with less influence on the effect deformation are filtered out by this method to reduce the complexity and diversity of the data, and at the same time reduce the number of samples required to establish the GA-SVR prediction model between the input variables and the objective function, thereby improving the prediction accuracy and generalization ability of the final model; and the value range of the required optimization parameter is selected according to the Spearman correlation coefficient of each characteristic geometric dimension;

[0021] Spearman correlation coefficient

[0022] Where dj is the difference between the variable-effect relationships, j is the number of critical geometric dimensions after screening, and k is the number of data points;

[0023] r s The value range of is from -1 to +1;

[0024] A positive correlation means that as one variable increases, the other variable tends to increase as well;

[0025] A negative correlation means that as one variable increases, the other tends to decrease;

[0026] r s Close to 0 means there is no significant monotonic relationship between the two. s Values ​​close to -1 or +1 indicate a strong correlation between the two;

[0027] The value range of the optimization parameter: L j ×(1+rs)>L , j >L j ×(1-rs);

[0028] L j is the original size of the jth key geometric dimension, L , j is the jth critical geometric dimension.

[0029] Furthermore, the Latin hypercube sampling method is used to generate approximate random samples within the range of multivariate parameter values, and the number of samples extracted by the Latin hypercube sampling is set, and all random samples are automatically batch simulated using a finite element simulation method;

[0030] Latin hypercube sampling is used to generate approximate random samples from the multivariate characteristic geometric parameter distribution determined in step 2. In this sampling method, the value range of each parameter is divided into several probability intervals, and a sample point is randomly selected in each interval. These randomly selected sample points are combined to form a final sample set. It ensures the diversity and representativeness of the samples by ensuring the uniform distribution of samples in multidimensional space, and is suitable for complex problems that require a large number of simulations and analyses. The number of samples required depends on multiple factors, including the complexity of the problem, the diversity of the data, the number of features, and the generalization ability of the model. If the complexity of the data set is high, or the data is widely distributed in the input space, more samples are needed to ensure that the model can capture all important features. After setting the required number of samples, the simulation samples can be submitted in the corresponding finite element simulation software and automatically batch simulated.

[0031] Furthermore, the objective function describing the forming quality is established, and the required data from all sample simulation results in step 3 are extracted as a data set for establishing a prediction model;

[0032] The input variables required by the data set in the above steps are the key geometric parameters that need to be optimized. The objective function of the data set is mainly to evaluate the sub-goals of the forging quality, including but not limited to: the effect variable average value y1 describing the deformation degree of the final forging, the effect variable standard deviation value y2 describing the deformation uniformity, the average grain size y3 describing the uniformity of the organization, the average grain size standard deviation value y4, and the recrystallization volume component y5. This involves the problem of multi-objective optimization. Now each sub-goal is normalized to obtain y , i ∈(0, 1), and then the normalized sub-goals are integrated using linear weighting to obtain the objective function M is the number of sub-goals selected, and the weight of each sub-goal is w l The allocation needs to consider the influence of the average value of each target effect, the standard deviation of the effect, the average grain size, the standard deviation of the average grain size, and the recrystallization volume component on the final forming quality of the forging. The greater the influence on the final forming quality of the forging, the greater the weight value. The weight of each sub-target satisfies Because the smaller the standard deviation of equivalent strain, average grain size, and standard deviation of average grain size, the better, positive numbers are taken, and the larger the average value of equivalent strain and recrystallization volume component, the better the negative numbers are taken. The smaller the final objective function Y, the better the forging quality. The corresponding data is automatically extracted from the file with simulation results in batches through the corresponding keywords through the python script. The extracted data includes the geometric size parameters of each sample and the effective strain and grain size of all grids at the end of the simulation. After processing the extracted data according to the required goals, the obtained data is finally used as the data set for establishing the prediction model. The data set is a multi-input and single-output table file.

[0033] Mean effect size

[0034] Effect variable standard deviation

[0035] Average grain size

[0036] Average grain size standard deviation

[0037] Recrystallized volume fraction

[0038] ∈i is the effect change of the ith grid, is the mean effect size, d i is the grain size of the ith grid, is the average grain size, is the recrystallized volume component of the ith grid, n is the total number of grids;

[0039] Min-max Normalization

[0040] Among them, y l is each sub-goal, y , l This method scales all data to the (0, 1) interval.

[0041] Further, the GA-SVR prediction model between the input variables and the objective function is established through the data set of step 4, the optimal value of the objective function is predicted, and the optimal combination of geometric parameters of the pre-forging is obtained;

[0042] Call the data set established in step 4, combine the support vector regression machine with the genetic algorithm, and build a prediction optimization model for the deformation uniformity of high-temperature alloy disk forgings;

[0043] Sample set for model building: (a1,b1), (a2,b2)...(a p ,bp )...(a t ,b t ), a p ,b p ∈R

[0044] The objective function of SVR can be expressed as:

[0045] Subject to the following conditions:

[0046] The kernel function of SVR is:

[0047] Among them, a p is the input variable value of the support vector regression machine, b p is the corresponding output variable value, φ(a p ) is the kernel function that maps data to a high-dimensional space, ω is the weight vector, μ is the bias term, ξ p and is a slack variable used to handle points outside the insensitive interval; C is a regularization parameter that controls the penalty intensity for interval violations; kernel function parameter γ and error tolerance ε;

[0048] The optimal solution of the problem is searched through selection, crossover and mutation operations in the iterative process of the genetic algorithm; when optimizing the SVR model, the genetic algorithm is used to optimize the parameters of the SVR, including the regularization parameter C, the kernel function parameter γ and the error tolerance ε. The range of each parameter is set to (0, 100), (0, 1000), (0, 10). The optimization process is divided into the following steps

[0049] Encoding: Encode the SVR parameters, regularization parameter C, kernel function parameter γ and error tolerance ε into chromosomes;

[0050] Initialize the population: randomly generate a set of solutions as the initial population;

[0051] Fitness calculation: Use cross-validation method to calculate the fitness of each individual, usually the prediction accuracy;

[0052] Selection, crossover, and mutation: The chromosomes in the population evolve toward the optimal values ​​of C, γ, and ε through selection, crossover, and mutation, and generate new populations;

[0053] Termination condition: Repeat the above process until the termination condition is met, the maximum number of iterations is reached, or the fitness reaches a certain threshold;

[0054] After the optimal model is obtained by iteratively optimizing the support vector regression model parameters using a genetic algorithm, the optimal value of the objective function is predicted to obtain the best combination of pre-forging geometric parameters.

[0055] Another object of the present invention is to provide a high-temperature alloy disc pre-forging shape optimization system comprising:

[0056] Design module, used to simulate the disc-like parts without pre-forging process. According to the overall effect deformation of the disc-like parts after final forging, the reverse simulation technology is used to quickly design the initial geometric shape of the pre-forging parts, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts;

[0057] A screening module is used to combine the experimental design method with the finite element method to conduct screening experimental design to determine the characteristic geometric parameters of the pre-forging as key optimization parameters and determine the value range of the optimization parameters;

[0058] The simulation module is used to determine the number of samples extracted by hypercube sampling and to automatically perform batch simulation on all random samples using the finite element simulation method;

[0059] An extraction module is used to establish an objective function describing the forming quality and extract the required data from all sample simulation results as a data set for establishing a prediction model;

[0060] The prediction module is used to establish a GA-SVR prediction model between input variables and objective functions through data sets, predict the optimal value of the objective function, and obtain the best combination of pre-forging geometric parameters.

[0061] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the high-temperature alloy disk pre-forging shape optimization method.

[0062] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the high-temperature alloy disc pre-forging shape optimization method.

[0063] Another object of the present invention is to provide an information data processing terminal, which is used to implement the high-temperature alloy disc pre-forging shape optimization system.

[0064] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0065] First, the present invention combines finite element numerical simulation, support vector machine and genetic algorithm to establish a GA-SVR prediction model between input variables and objective functions, predict the optimal value of the objective function, and obtain the best combination of geometric parameters of pre-forgings. First, the shape of the pre-forgings is preliminarily designed through experience, and then the results of finite element simulation are extracted to establish a data set, and the data set is used to establish the parameters of the support vector machine model as the objective function of the genetic algorithm, and the parameters of the support vector machine model are optimized using the genetic algorithm to obtain the best prediction model. The size parameter combination of the pre-forgings when the deformation uniformity evaluation index is the best is obtained. The global optimization of the genetic algorithm effectively improves the deformation uniformity of the final forging, thereby improving the final performance of the forgings. Compared with the traditional pre-forging design method, this method has the advantages of strong versatility, less manual operation, and good deformation uniformity, shortens the design cycle, improves production efficiency, and reduces production costs.

[0066] Second, the technical solution of the present invention fills the technical gap in the industry at home and abroad: it provides an efficient method and system for optimizing the shape of high-temperature alloy disc pre-forgings, automatically selects samples and simulates in batches through the latest finite element simulation software, greatly improves the efficiency in experimental scheme design and simulation result acquisition, and then extracts the simulation results through the script to produce a data set, and optimizes the shape of the pre-forging by establishing an intelligent algorithm model. This method is not only suitable for high-temperature alloy disc parts, but also suitable for parts that cannot be mass-produced in actual production and need to be continuously sought for the optimal design through simulation. It only needs to use the same method to obtain batches of samples, then extract the simulation results to establish a data set, and then establish an intelligent algorithm model of design variables and objective functions according to the situation, and then the intelligent algorithm model automatically optimizes the design variables.

[0067] The technical solution of the present invention overcomes technical bias: an intelligent algorithm model is established through a data set, and then optimization is performed. This method combines the advantages of machine learning. The intelligent algorithm can automate the optimization process, reduce the cost and time of manual trial and error, and improve the optimization efficiency; the model established based on the data set can provide data-driven decision support, increase the objectivity and accuracy of decision-making; many intelligent optimization algorithms, such as genetic algorithms, particle swarm optimization, etc., have global optimization capabilities, can avoid local optimal solutions, and increase the probability of finding the global optimal solution; optimization through intelligent algorithm models can reduce the need for physical experiments, thereby reducing costs and risks; the optimization process of the intelligent algorithm model can be automated, and the optimization strategy can be flexibly adjusted according to different business needs. This method can be extended to large-scale data sets and high-dimensional data, and is suitable for problems of various sizes.

[0068] Third, the shape optimization method of high-temperature alloy disc pre-forgings provided by the present invention significantly solves many problems in the prior art in industrial applications and achieves technological progress in many aspects.

[0069] 1. Solve the problems of uneven deformation and coarse grains of forgings in the prior art: Turbine disk forgings under traditional forging processes are difficult to meet the requirements of high precision and high performance, especially the problems of uneven deformation and coarse grains of forgings. Through the reverse simulation method of the present invention, the shape of the pre-forging can be designed more scientifically, so that the equivalent strain distribution of the forging after final forging is more uniform, which significantly improves the deformation uniformity of the forging, thereby improving the microstructure quality of the forging, refining the grain structure, and meeting the stringent requirements of key components of aircraft engines.

[0070] 2. Improve optimization efficiency and reduce production costs and time: Traditional empirical analog design and physical experimental verification methods are time-consuming and labor-intensive, and are difficult to cope with complex forging design. The experimental design method of the present invention is combined with finite element simulation to screen out the geometric parameters that have the greatest impact on forging performance, and simulated samples are generated through Latin hypercube sampling, which reduces the number of physical experiments and the amount of data required, thereby accelerating the development process. This method reduces the need for repeated experiments, shortens the R&D cycle, and reduces production costs. It is particularly suitable for the design needs of pre-forgings with multiple parameters and complex structures.

[0071] 3. Provide accurate multi-objective optimization design: The existing response surface model is not good enough in dealing with complex nonlinear relationships and it is difficult to effectively capture the deep relationship between the design variables. The present invention uses a multi-objective optimization method to normalize multiple sub-objectives that affect the forging quality (such as equivalent strain, grain size and recrystallization volume component), and constructs an overall objective function based on a linear weighted method, providing a more reliable evaluation standard for the forging quality, thereby ensuring the accuracy of the optimization design and the forging quality.

[0072] 4. Improvement of intelligence and automation: The present invention adopts support vector regression (SVR) combined with genetic algorithm (GA) for multi-objective optimization modeling, realizing the intelligence and automation of the optimization process. The parameters of the SVR model are automatically optimized by genetic algorithm, so that the model has higher prediction accuracy and generalization ability. The optimized model can efficiently predict the best combination of pre-forging geometric parameters, improve the forming quality of forgings, avoid frequent manual intervention in traditional optimization methods, and improve optimization efficiency.

[0073] In summary, the method of the present invention effectively solves the problems of uneven deformation and coarse grains of forgings through precise pre-forging shape optimization and intelligent, multi-objective optimization modeling in industrial applications, improves the forming quality and production efficiency of forgings, achieves technological progress in forging shape optimization, and provides a strong guarantee for the reliability and safety of high-temperature alloy disc forgings. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a flow chart of a method for optimizing the shape of a high-temperature alloy disc pre-forging provided in an embodiment of the present invention.

[0075] Figure 2 It is a structural block diagram of a high-temperature alloy disc pre-forging shape optimization system provided by an embodiment of the present invention.

[0076] Figure 3 It is a diagram of a point tracking process performed through finite element simulation software post-processing provided by an embodiment of the present invention.

[0077] Figure 4 It is the initial pre-forging shape and key geometric parameter diagram provided by the embodiment of the present invention.

[0078] Figure 5 It is a relationship diagram of the influence of key geometric parameters on equivalent strain provided by an embodiment of the present invention.

[0079] Figure 6 It is a schematic diagram of SVR provided by an embodiment of the present invention.

[0080] Figure 7 The equivalent strain distribution cloud map, grain size distribution cloud map, and recrystallization volume component cloud map before and after optimization provided by the embodiments of the present invention;

[0081] Figure 8 is the fitting error between the predicted value and the actual value of the GA-SVR model training sample provided by the embodiment of the present invention;

[0082] Fig. 9 It is the fitting error between the predicted value and the actual value of the GA-SVR model test sample provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0084] The following are two specific embodiments of the present invention applied in actual industry:

[0085] Example 1: Optimization design of aircraft engine turbine disk preforgings

[0086] In the manufacture of aircraft engines, turbine disks are key high-temperature alloy disk forgings, and their performance directly affects the reliability of the engine. Using the optimization method of the present invention, the optimization design of turbine disk pre-forgings can be achieved through the following steps:

[0087] 1. Initial design by reverse simulation: First, the final forging simulation without pre-forging process is carried out on the turbine disk to obtain its deformation distribution and equivalent strain distribution, and the initial geometric shape of the pre-forging is designed by reverse simulation technology.

[0088] 2. Key parameter screening and optimization: The experimental design method is combined with finite element simulation to extract the key geometric parameters that affect the final forging deformation uniformity and grain refinement, and obtain the pre-forging geometric dimensions most suitable for the final forging of the turbine disk.

[0089] 3. Multi-objective optimization: The Latin hypercube sampling method is used to generate samples and carry out simulations. The equivalent strain, grain size distribution, recrystallization volume component and other indicators of the forgings are calculated. An optimization model is established based on GA-SVR, and the optimal value of the objective function is predicted to obtain the final pre-forging shape.

[0090] Through this optimization process, the final forging of the turbine disk has achieved a dense structure, uniform distribution of refined grains, significantly improved mechanical properties, and met the strict quality standards of key aircraft engine components. At the same time, compared with traditional methods, the R&D cycle is shortened by about 30% and the production cost is reduced by more than 20%.

[0091] Example 2: Precision Forging Optimization of Industrial Gas Turbine Discs

[0092] In industrial gas turbines, the wheel disc needs to withstand extreme environments of high temperature and high stress for a long time, and the quality of its forgings directly determines the service life and performance stability of the turbine. The specific process of the optimization method of the present invention for the design of wheel disc pre-forgings is as follows:

[0093] 1. Final forging simulation and pre-forging design: The wheel disc is simulated by finite element method for final forging to obtain the equivalent strain and grain distribution results. The inverse simulation method is used to quickly design the initial pre-forging shape with a reasonable volume distribution to improve the deformation uniformity after final forging.

[0094] 2. Multi-parameter optimization experimental design: Determine the key geometric parameters and use Latin hypercube sampling to generate samples with different parameter combinations. Perform simulation calculations on each group of samples, analyze key quality indicators such as deformation uniformity and recrystallization volume component, and establish a GA-SVR prediction model that includes key geometric parameters and forming quality targets.

[0095] 3. Optimal shape optimization: Through iterative optimization of genetic algorithms, the optimal combination of geometric parameters is obtained to ensure that the mechanical properties of the wheel forging meet the design requirements and reduce the adjustment of subsequent processing and heat treatment.

[0096] After adopting this optimized design, the final forging forming accuracy of the gas turbine wheel disc was improved by about 15%, and the grain uniformity was greatly improved, meeting the service requirements under complex working conditions.

[0097] like Figure 1 As shown, a method for optimizing the shape of a high-temperature alloy disc pre-forging provided by an embodiment of the present invention comprises the following steps:

[0098] S101: Simulate the disc-like parts without pre-forging process. According to the overall equivalent strain of the disc-like parts after final forging, use the reverse simulation technology to quickly design the initial geometric shape of the pre-forging parts, and design the shape of the pre-forging die according to the shape of the pre-forging parts.

[0099] The final forging finite element simulation is directly performed on the billet. The equivalent strain results of the forging obtained by the finite element simulation are used to reversely trace the positions of the small deformation zone and the large deformation zone in the original billet, and the shape of the pre-forging is preliminarily designed. This method can eliminate the small deformation zone and the large deformation zone, which can significantly improve the deformation uniformity of the forging. The specific design method is to first divide it into N rectangular areas of equal volume along the radial direction according to the deformation after the final forging. If the shape of the final forging is more complex, the number of partitions for the final forging will also be more. Then, according to the partitions for the final forging, the billet is divided into N rectangular areas of the same volume. Write a script to extract the strain value ∈i of each grid after the final forging, and calculate the average equivalent strain value of the forging after the final forging. Then, the deformation conditions of the large deformation area and the small deformation area of ​​the final forging are analyzed, and the small deformation area and the large deformation area smaller than the forging are marked. The reverse tracking method is used to track the positions of the small deformation area and the large deformation area in the original billet. The points that need to be tracked after final forging are selected by point tracking in the post-processing process of finite element simulation software, and the positions of the selected points in the billet are tracked by reverse simulation. Then, the material in the billet and its corresponding small deformation area is reduced, and the reduced volume is added to the large deformation area along the radial direction. Finally, the sharp parts of each different rectangular area are smoothed to obtain the initial pre-forging shape.

[0100] Average equivalent strain value

[0101] ∈i is the equivalent strain of the ith grid, n is the total number of grids;

[0102]

[0103] α and β are parameters for selecting the overall deformation, and their values ​​range from 0.1 to 0.5.

[0104] S102: By combining the experimental design method with the finite element method, a screening experimental design is carried out to determine the characteristic geometric parameters of the S101 pre-forging as the key optimization parameters, and to determine the value range of the optimization parameters.

[0105] The geometric dimensions that mainly affect the equivalent strain of the forging during the pre-forging process are selected as the key geometric dimensions. The selection method is to extract all the geometric parameters of the pre-forging die designed in step 1, and then set all the parameters as the design variables of the finite element simulation. The experimental design method is combined with the finite element method, and a screening experimental design is performed to evaluate the influence of the geometric dimensions of the die on the equivalent strain of the forging. In the post-processing of the finite element simulation, the relationship diagram of the influence of all geometric parameters on the equivalent strain is obtained. The Spearman correlation coefficient (rs) is used to evaluate the influence of each geometric parameter variable on the equivalent strain. When the Spearman correlation coefficient is greater than q, the parameter is selected as the characteristic geometric parameter. q is the threshold value of the geometric parameter having a greater influence on the equivalent strain, ranging from 0.1 to 0.3. This method is used to filter out the geometric dimensions with less influence on the equivalent strain, so as to reduce the complexity and diversity of the data, and at the same time reduce the number of samples required to establish the GA-SVR prediction model between the input variables and the objective function, thereby improving the prediction accuracy and generalization ability of the final model. And the value range of the required optimization parameter is selected according to the Spearman correlation coefficient of each characteristic geometric dimension.

[0106] Spearman correlation coefficient

[0107] Where dj is the rank difference between the variable-equivalent strain relationship, j is the number of critical geometric dimensions after screening, and k is the number of data points;

[0108] r s The value range of is from -1 to +1;

[0109] A positive correlation means that as one variable increases, the other variable tends to increase as well;

[0110] A negative correlation means that as one variable increases, the other tends to decrease;

[0111] r s Close to 0 means there is no significant monotonic relationship between the two. s Values ​​close to -1 or +1 indicate a strong correlation between the two;

[0112] The value range of the optimization parameter: L j ×(1+rs)>L , j >L j ×(1-rs);

[0113] Lj is the original size of the jth key geometric dimension, L , j is the jth critical geometric dimension.

[0114] S103: using the Latin hypercube sampling method to generate approximate random samples within the range of multivariate parameter values, setting the number of samples drawn by the Latin hypercube sampling, and using the finite element simulation method to automatically perform batch simulation on all random samples.

[0115] Latin hypercube sampling is used to generate approximate random samples from the multivariate characteristic geometric parameter distribution determined in S102. In this sampling method, the value range of each parameter is divided into several intervals of equal probability, and a sample point is randomly selected in each interval. These randomly selected sample points are combined to form the final sample set. It ensures the diversity and representativeness of the samples by ensuring the uniform distribution of samples in multidimensional space, and is suitable for complex problems that require a large number of simulations and analyses. The number of samples required depends on multiple factors, including the complexity of the problem, the diversity of the data, the number of features, and the generalization ability of the model. If the complexity of the data set is high, or the data is widely distributed in the input space, more samples are needed to ensure that the model can capture all important features. After setting the required number of samples, the simulation samples can be submitted in the corresponding finite element simulation software and automatically batch simulated.

[0116] S104: Establish an objective function that describes the forming quality, and extract the required data from all sample simulation results in S103 as a data set for establishing a prediction model.

[0117] The input variables required by the data set in the above steps are the key geometric parameters that need to be optimized. The objective function of the data set is mainly to evaluate the sub-goals of the forging quality, including but not limited to: the average value of the equivalent strain y1 describing the deformation degree of the final forging, the standard deviation of the equivalent strain y2 describing the uniform deformation, the average grain size y3 describing the uniformity of the organization, the standard deviation of the average grain size y4, the recrystallization volume component y5, etc., involving multi-objective optimization problems. Now each sub-goal is normalized to obtain y , i ∈(0, 1), and then the normalized sub-goals are integrated using linear weighting to obtain the objective function M is the number of sub-goals selected, and the weight of each sub-goal is w l The allocation needs to consider the influence of the average value of equivalent strain, standard deviation of equivalent strain, average grain size, standard deviation of average grain size, recrystallization volume component, etc. on the final forming quality of forgings. The greater the influence on the final forming quality of forgings, the greater the weight value. The weight of each sub-target satisfies Because the smaller the standard deviation of equivalent strain, average grain size, and standard deviation of average grain size, the better, positive numbers are taken. The larger the average value of equivalent strain and recrystallization volume component, the better the negative numbers are taken. The smaller the final objective function Y, the better the forging quality. The corresponding data is automatically extracted from the file with simulation results in batches by the corresponding keywords through the python script. The extracted data includes the geometric size parameters of each sample and the equivalent strain and grain size of all grids at the end of the simulation. After processing the extracted data according to the required goals, the obtained data is finally used as the data set for establishing the prediction model. The data set is a multi-input and single-output table file.

[0118] Average equivalent strain value

[0119] Standard deviation of equivalent strain

[0120] Average grain size

[0121] Average grain size standard deviation

[0122] Recrystallized volume fraction

[0123] ∈i is the equivalent strain of the ith grid, is the average equivalent strain value, d i is the grain size of the ith grid, is the average grain size, is the recrystallized volume component of the ith grid, n is the total number of grids;

[0124] Min-max Normalization

[0125] Among them, y l is each sub-goal, y , l This method scales all data to the (0, 1) interval.

[0126] S105: A GA-SVR prediction model between input variables and objective functions is established through the data set of S104, and the optimal value of the objective function is predicted to obtain the best combination of geometric parameters of the pre-forging.

[0127] The data set established in S104 is called, and the support vector regression machine is combined with the genetic algorithm to construct a prediction optimization model for the deformation uniformity of high-temperature alloy disk forgings.

[0128] Sample set for model building: (a1,b1), (a2,b2)...(a p,b p )...(a t ,b t ), a p ,b p ∈R

[0129] The objective function of SVR can be expressed as:

[0130] Subject to the following conditions:

[0131] The kernel function of SVR is:

[0132] Among them, a p is the input variable value of the support vector regression machine, b p is the corresponding output variable value, φ(a p ) is the kernel function that maps data to a high-dimensional space, ω is the weight vector, μ is the bias term, ξ p and is a slack variable used to deal with points outside the insensitive interval; C is a regularization parameter that controls the penalty intensity for interval violations; kernel function parameter γ and error tolerance ε.

[0133] The optimal solution to the problem is searched through selection, crossover and mutation operations in the iterative process of the genetic algorithm. When optimizing the SVR model, the genetic algorithm is used to optimize the parameters of the SVR, including the regularization parameter C, the kernel function parameter γ and the error tolerance ε. The range of each parameter is set to (0, 100), (0, 1000), (0, 10). The optimization process is divided into the following steps

[0134] Encoding: Encode the SVR parameters, regularization parameter C, kernel function parameter γ and error tolerance ε into chromosomes;

[0135] Initialize the population: randomly generate a set of solutions as the initial population;

[0136] Fitness calculation: Use methods such as cross-validation to calculate the fitness of each individual, usually the prediction accuracy;

[0137] Selection, crossover, and mutation: The chromosomes in the population evolve toward the optimal values ​​of C, γ, and ε through selection, crossover, and mutation, and generate new populations;

[0138] Termination condition: Repeat the above process until the termination condition is met, the maximum number of iterations is reached, or the fitness reaches a certain threshold;

[0139] After the optimal model is obtained by iteratively optimizing the support vector regression model parameters using a genetic algorithm, the optimal value of the objective function is predicted to obtain the best combination of pre-forging geometric parameters.

[0140] like Figure 2 As shown, the high-temperature alloy disc pre-forging shape optimization system provided by the embodiment of the present invention includes:

[0141] Design module, used to simulate the disc-like parts without pre-forging process. According to the overall effect deformation of the disc-like parts after final forging, the reverse simulation technology is used to quickly design the initial geometric shape of the pre-forging parts, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts;

[0142] A screening module is used to combine the experimental design method with the finite element method to conduct screening experimental design to determine the characteristic geometric parameters of the pre-forging as key optimization parameters and determine the value range of the optimization parameters;

[0143] The simulation module is used to determine the number of samples extracted by hypercube sampling and to automatically perform batch simulation on all random samples using the finite element simulation method;

[0144] An extraction module is used to establish an objective function describing the forming quality and extract the required data from all sample simulation results as a data set for establishing a prediction model;

[0145] The prediction module is used to establish a GA-SVR prediction model between input variables and objective functions through data sets, predict the optimal value of the objective function, and obtain the best combination of pre-forging geometric parameters.

[0146] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the high-temperature alloy disk pre-forging shape optimization method.

[0147] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the high-temperature alloy disc pre-forging shape optimization method.

[0148] Another object of the present invention is to provide an information data processing terminal, which is used to implement the high-temperature alloy disc pre-forging shape optimization system.

[0149] The present invention is specifically implemented:

[0150] (1) The disc-like parts are simulated without pre-forging process. According to the overall equivalent strain of the disc-like parts after final forging, the initial geometric shape of the pre-forging parts is quickly designed using reverse simulation technology, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts.

[0151] Finite element simulation software is used to simulate the roughening and final forging of the cylindrical billet. According to the overall equivalent strain after final forging, the billet before final forging is divided into five regions of equal volume along the radial direction. A script is written to extract the strain value ∈i of each grid after final forging, and then the average equivalent strain value of the forging after final breaking is calculated. is 2.54, α and β are 0.3 and 0.5, and the small deformation area ∈ small ≤ 1.78, and the large deformation area ∈ large ≥ 3.81. Then the reverse tracking method is used to track the positions of the small deformation area and the large deformation area in the original billet. The points to be tracked after final forging are selected by point tracking in the post-processing process of the finite element simulation software, and then the reverse simulation is used to track the positions of the selected points in the billet. The tracking process is as follows Figure 3 As shown, P1-P30 are points in the small deformation area, P31-P50 are points in the large deformation area, and then the material of the cake blank and its corresponding small deformation area is reduced, and the reduced volume is added to the large deformation area along the radial direction. Then, the sharp part of each different rectangular area is smoothed to obtain the initial pre-forging shape as shown in Figure 4 As shown, a pre-forging die is designed according to the shape of the pre-forging.

[0152] (2) By combining the experimental design method with the finite element method, a screening experimental design is carried out to determine that the key geometric parameters of the pre-forging in step 1 are the required optimization parameters, and the value range of the required optimization parameters is determined.

[0153] A multi-process simulation of high-temperature alloy disc forgings was established through finite element simulation software. The process flow includes blanking-heating-roughing-heating-pre-forging-heating-final forging. After the simulation of the whole process, all geometric parameters of the pre-forging process were extracted for screening experimental design to evaluate the influence of these geometric dimension changes on the equivalent effect strain. The relationship diagram of the influence of all geometric parameters on the equivalent effect strain was obtained in the post-processing of the finite element simulation. Figure 5 As shown in the figure, the key geometric parameters are screened out through the Spearman correlation coefficient in the figure. When the Spearman correlation coefficient is greater than or equal to 0.2, the parameter is selected as the key geometric parameter. The key geometric size parameters of the pre-forging die are added in the pre-forging process as input variables. The three variables of the upper and lower dies are obtained through screening as follows Figure 4 As shown, the formula L is calculated according to the value range of the optimization parameter j ×(1±rs), the value ranges are as follows: lower die H1 is 35~65mm, lower die H2 is 20~40mm, lower die D1 is 15~45mm, upper die H3 is 17~47mm, upper die H4 is 8~28mm, and upper die D2 is 40~80mm.

[0154] (3) The Latin hypercube sampling method is used to generate approximate random samples within the range of multivariate parameter values, and the number of samples drawn by the Latin hypercube sampling is set. The finite element simulation software is used to automatically batch simulate all random samples.

[0155] Latin hypercube sampling is used to generate approximate random samples from the multivariate key geometric parameter distribution determined in step 2. The input variables include six geometric dimensions. The number of samples is set to 50, including 40 training set samples and 10 test set samples. After setting the required number of samples, the simulation samples can be submitted to the corresponding finite element simulation software and automatically batch simulated.

[0156] (4) Establish an objective function that describes the forming quality, and extract the required data from all sample simulation results in step 3 as the data set for establishing the prediction model.

[0157] The input variables of the data set are the key geometric parameters that need to be optimized. The objective function of the data set describes the deformation degree of the final forging, the equivalent strain average value y1, the equivalent strain standard deviation value y2, the average grain size y3, the average grain size standard deviation y4, and the recrystallization volume component y5. It involves a multi-objective optimization problem. Now each sub-goal is normalized to obtain y , l ∈(0,1), and then the normalized sub-goals are integrated using linear weighting to obtain the objective function The weight w of each sub-goal l The allocation needs to consider the influence of the average value of equivalent strain, standard deviation of equivalent strain, average grain size, standard deviation of average grain size, recrystallization volume component, etc. on the final forming quality of forgings. The greater the influence on the final forming quality of forgings, the greater the weight value. The weight of each sub-goal satisfies Because the smaller the standard deviation of equivalent strain, average grain size, and standard deviation of average grain size, the better, positive numbers are taken. The larger the average value of equivalent strain and recrystallization volume component, the better the negative numbers are taken. The smaller the final objective function Y, the better the forging quality. The corresponding data is automatically extracted from the file with simulation results in batches by the corresponding keywords through the python script. The extracted data includes the geometric size parameters of each sample and the equivalent strain and grain size of all grids at the end of the simulation. After processing the extracted data according to the required goals, the obtained data is finally used as the data set for establishing the prediction model. The data set is a multi-input and single-output table file.

[0158] (5) Through the data set of step 4, a GA-SVR prediction model between the input variables and the objective function is established to predict the optimal value of the objective function and obtain the best combination of pre-forging geometric parameters.

[0159] The data set established in step 4 is called, and the support vector regression machine is combined with the genetic algorithm to build a prediction optimization model for the deformation uniformity of high-temperature alloy disc forgings. The 40 samples in the data set are used as the training set for the SVR model, and the 10 samples are used as the test set. The SVR schematic diagram is shown in the figure. Figure 6 As shown. A GA-SVR prediction model is constructed using a genetic algorithm. The genetic algorithm is used to optimize the parameters of SVR, the regularization parameter C, the kernel function parameter γ and the error tolerance ε. The range of each parameter is set to (0,100), (0,1000), (0,10). The setting of parameters in the genetic algorithm directly affects the optimization speed and quality of the algorithm. The parameter settings are determined as follows: the number of initialized populations is 200, the maximum genetic generation is 200, the random traversal sampling function sus is selected, the two-point crossover probability is 0.8, and the discrete mutation probability is 0.007. Finally, the optimal value of the objective function is predicted to obtain the best combination of pre-forging geometric parameters.

[0160] The optimal value of the objective function predicted by the GA-SVR model is Y=3.05, and the corresponding pre-forging geometric parameter combination is lower die H1 of 60.64mm, lower die H2 of 24.47mm, lower die D1 of 23.88mm, upper die H3 of 22.93mm, upper die H4 of 11.71mm, and upper die D2 of 49.81mm.

[0161] The pre-forging part is drawn in CAD software, and the corresponding pre-forging die is designed. Finally, the solution is simulated and verified in finite element simulation software. The simulation results after optimization are compared with those before optimization. The equivalent strain distribution cloud map and grain size distribution cloud map before and after optimization are shown in Figure 2. Figure 7 shown.

[0162] The above results indicate that the method proposed in the present invention can effectively carry out multi-objective optimization design of the shape of pre-forgings of high-temperature alloy disk-like parts. The shape of the pre-forgings designed by the method can meet the requirements of uniform deformation, grain refinement and sufficient recrystallization of high-temperature alloy disk-like parts while ensuring complete filling, thereby providing a feasible method for improving the quality of high-temperature alloy disk-like parts.

[0163] The present invention can be applied to rotating core components of aerospace engines formed by forging technology. These core components work in extremely harsh environments for a long time, and the comprehensive performance requirements of such forgings after forming are relatively strict. Related products mainly include fan disks, compressor disks, turbine disks, shaft necks and rotor support cone walls.

[0164] The geometric parameter combination of each sample under Latin hypercube sampling and the corresponding objective function are shown in Table 1. Therefore, the average value of Y in the sample is 3.36, the minimum objective function of sample 6 is Y=3.14, and the corresponding pre-forging geometric parameter combination is lower die H1 is 49.78mm, lower die H2 is 20.81mm, lower die D1 is 31.20mm, upper die H3 is 22.93mm, upper die H4 is 13.32mm, and upper die D2 is 75.25mm.

[0165]

[0166]

[0167] Table 1

[0168] like Figure 8 and Fig. 9 As shown in the figure, the GA-SVR model prediction is established using the above data set. The mean square error between the predicted value and the actual value of the objective function of the training sample is 0.13161, and the mean square error between the predicted value and the actual value of the objective function of the test sample is 0.1575. The GA-SVR model has a high degree of fit between the predicted pre-forging geometric parameters and the objective function.

[0169] The prediction model is applied to all schemes within the range of pre-forging geometric parameters, and the optimal value of the objective function is predicted. The optimal value of the objective function Y=3.07 predicted by the GA-SVR model and the corresponding pre-forging geometric parameter combination are shown in Table 2 below.

[0170]

[0171] Table 2

[0172] The comparison of each sub-objective function using the optimal sample of Latin hypercube sampling objective function and the optimal prediction geometric parameter combination of the objective function predicted by the GA-SVR model is shown in Table 3.

[0173]

[0174] Table 3

[0175] By comparing the objective function and sub-objective functions before and after optimization by this method, we can know that the average value of Y in the sample is 3.36, the minimum Y value in the sample is 3.14, and the Y value after GA-SVR model optimization is 3.07, which is a significant improvement compared with the sample, and each sub-objective function is basically developing in the direction of improving the forming quality of forgings.

[0176] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0177] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for optimizing the shape of a high-temperature alloy disc pre-forging, characterized in that: The following steps are involved: Step 1: Simulate the disc-like parts without pre-forging process. According to the overall equivalent strain of the disc-like parts after final forging, use the reverse simulation technology to quickly design the initial geometric shape of the pre-forging parts, and design the shape of the pre-forging die according to the shape of the pre-forging parts; Step 2: By combining the experimental design method with the finite element method, a screening experimental design is performed to determine the characteristic geometric parameters of the pre-forging in step 1 as the key optimization parameters, and determine the value range of the optimization parameters; Step 3: Use the Latin hypercube sampling method to generate approximate random samples within the range of multivariate parameter values, set the number of samples drawn by the Latin hypercube sampling, and use the finite element simulation method to automatically batch simulate all random samples; Step 4: Establish an objective function that describes the forming quality, and extract the required data from all sample simulation results in step 3 as a data set for establishing a prediction model; Step 5: A GA-SVR prediction model between the input variables and the objective function is established through the data set in step 4, the optimal value of the objective function is predicted, and the best combination of pre-forging geometric parameters is obtained.

2. The method for optimizing the shape of a high-temperature alloy disc pre-forging according to claim 1, characterized in that: The simulation of the disc-like parts without pre-forging process is carried out, and according to the overall effect deformation of the disc-like parts after final forging, the initial geometric shape of the pre-forging parts is quickly designed by using the reverse simulation technology, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts: The final forging finite element simulation is directly performed on the billet. The forging effect deformation results obtained by the finite element simulation are used to reversely track the positions of the small deformation zone and the large deformation zone in the original billet, and the shape of the pre-forging is preliminarily designed. This method can eliminate the small deformation zone and the large deformation zone, and can significantly improve the deformation uniformity of the forging. The specific design method is to first divide the billet into N rectangular areas of volume phase along the radial direction according to the deformation after the final forging. If the shape of the final forging is more complex, the number of partitions for the final forging is also more. Then, according to the partitions of the final forging, the billet is divided into N rectangular areas of the same volume. Write a script to extract the strain value ∈i of each grid after final forging, and calculate the average effect strain value of the forging after final forging Then, the deformation of the large deformation area and the small deformation area of ​​the final forging is analyzed, and the small deformation area and the large deformation area smaller than the forging are marked, and the reverse tracking method is used to track the positions of the small deformation area and the large deformation area in the original billet. The points that need to be tracked after the final forging are selected by point tracking in the post-processing process of the finite element simulation software, and the positions of the selected points in the billet are tracked by reverse simulation. Then, the material of the billet and its corresponding small deformation area is reduced, and the reduced volume is added to the large deformation area along the radial direction. Finally, the sharp part of each different rectangular area is smoothed to obtain the initial pre-forging shape; Mean effect size ∈i is the effect strain of the ith grid, n is the total number of grids; α and β are parameters for selecting the overall deformation, and their values ​​range from 0.1 to 0.

5.

3. The method for optimizing the shape of a high-temperature alloy disc pre-forging according to claim 1, characterized in that: The experimental design method is combined with the finite element method to perform a screening experimental design to determine the characteristic geometric parameters of the pre-forging in step 1 as key optimization parameters, and determine the value range of the optimization parameters; The geometric dimensions that mainly affect the effect deformation of the forging during the pre-forging process are selected as the key geometric dimensions. The selection method is to extract all the geometric parameters of the pre-forging die preliminarily designed in step 1, and then set all the parameters as design variables of the finite element simulation. The experimental design method is combined with the finite element method, and a screening experimental design is performed to evaluate the influence of the geometric dimensions of the die on the effect deformation of the forging. In the post-processing of the finite element simulation, a relationship diagram of the influence of all geometric parameters on the effect deformation is obtained. The Spearman correlation coefficient (rs) is used to evaluate the influence of each geometric parameter variable on the effect deformation. When the Spearman correlation coefficient is greater than q, the parameter is selected as the characteristic geometric parameter. q is the threshold value of the geometric parameter having a greater influence on the effect deformation, ranging from 0.1 to 0.

3. The geometric dimensions with less influence on the effect deformation are filtered out by this method to reduce the complexity and diversity of the data, and at the same time reduce the number of samples required to establish the GA-SVR prediction model between the input variables and the objective function, thereby improving the prediction accuracy and generalization ability of the final model; and the value range of the required optimization parameter is selected according to the Spearman correlation coefficient of each characteristic geometric dimension; Spearman correlation coefficient Where dj is the difference between the variable-effect relationships, j is the number of critical geometric dimensions after screening, and k is the number of data points; r s The value range of is from -1 to +1; A positive correlation means that as one variable increases, the other variable tends to increase as well; A negative correlation means that as one variable increases, the other tends to decrease; r s Close to 0 means there is no significant monotonic relationship between the two. s Values ​​close to -1 or +1 indicate a strong correlation between the two; The value range of the optimization parameter: L j ×(1+rs)>L , j >L j ×(1-rs); L j is the original size of the jth key geometric dimension, L , j is the jth critical geometric dimension.

4. The method for optimizing the shape of a high-temperature alloy disc pre-forging according to claim 1, characterized in that: The Latin hypercube sampling method is used to generate approximate random samples within the range of multivariate parameter values, and the number of samples extracted by the Latin hypercube sampling is set, and all random samples are automatically batch simulated using a finite element simulation method; Latin hypercube sampling is used to generate approximate random samples from the multivariate feature geometric parameter distribution determined in step 2. In this sampling method, the value range of each parameter is divided into several probability intervals, and a sample point is randomly selected in each interval. These randomly selected sample points are combined to form the final sample set. It ensures the diversity and representativeness of the samples by ensuring the uniform distribution of samples in multidimensional space. It is suitable for complex problems that require a large number of simulations and analyses. The number of samples required depends on multiple factors, including the complexity of the problem, the diversity of the data, the number of features, and the generalization ability of the model. If the complexity of the data set is high, or the data is widely distributed in the input space, more samples are needed to ensure that the model can capture all important features. After setting the required number of samples, the simulation samples can be submitted to the corresponding finite element simulation software and automatically simulated in batches.

5. The method for optimizing the shape of a high-temperature alloy disc pre-forging according to claim 1, characterized in that: The objective function describing the forming quality is established, and the required data from all sample simulation results in step 3 are extracted as a data set for establishing a prediction model; The input variables required by the data set in the above steps are the key geometric parameters that need to be optimized. The objective function of the data set is mainly to evaluate the sub-goals of the forging quality, including but not limited to: the effect variable average value y1 describing the deformation degree of the final forging, the effect variable standard deviation value y2 describing the deformation uniformity, the average grain size y3 describing the uniformity of the organization, the average grain size standard deviation value y4, and the recrystallization volume component y5. This involves the problem of multi-objective optimization. Now each sub-goal is normalized to obtain y , i ∈(0, 1), and then the normalized sub-goals are integrated using linear weighting to obtain the objective function M is the number of sub-goals selected, and the weight of each sub-goal is w l The allocation needs to consider the influence of the average value of each target effect, the standard deviation of the effect, the average grain size, the standard deviation of the average grain size, and the recrystallization volume component on the final forming quality of the forging. The greater the influence on the final forming quality of the forging, the greater the weight value. The weight of each sub-target satisfies Because the smaller the standard deviation of equivalent strain, average grain size, and standard deviation of average grain size, the better, positive numbers are taken, and the larger the average value of equivalent strain and recrystallization volume component, the better the negative numbers are taken. The smaller the final objective function Y, the better the forging quality. The corresponding data is automatically extracted from the file with simulation results in batches through the corresponding keywords through the python script. The extracted data includes the geometric size parameters of each sample and the effective strain and grain size of all grids at the end of the simulation. After processing the extracted data according to the required goals, the obtained data is finally used as the data set for establishing the prediction model. The data set is a multi-input and single-output table file. Mean effect size Effect variable standard deviation Average grain size Average grain size standard deviation Recrystallized volume fraction ∈i is the effect change of the ith grid, is the mean effect size, d i is the grain size of the ith grid, is the average grain size, is the recrystallized volume component of the ith grid, n is the total number of grids; Min-max Normalization Among them, y l is each sub-goal, y , l This method scales all data to the (0, 1) interval.

6. The method for optimizing the shape of a high-temperature alloy disc pre-forging according to claim 1, characterized in that: The GA-SVR prediction model between the input variables and the objective function is established through the data set of step 4, the optimal value of the objective function is predicted, and the optimal combination of geometric parameters of the pre-forging is obtained; Call the data set established in step 4, combine the support vector regression machine with the genetic algorithm, and build a prediction optimization model for the deformation uniformity of high-temperature alloy disk forgings; Sample set for model building: (a1,b1), (a2,b2)...(a p ,b p )...(a t ,b t ), a p ,b p ∈R The objective function of SVR can be expressed as: Subject to the following conditions: The kernel function of SVR is: Among them, a p is the input variable value of the support vector regression machine, b p is the corresponding output variable value, φ(a p ) is the kernel function that maps data to a high-dimensional space, ω is the weight vector, μ is the bias term, ξ p and is a slack variable used to handle points outside the insensitive interval; C is a regularization parameter that controls the penalty intensity for interval violations; kernel function parameter γ and error tolerance ε; The optimal solution of the problem is searched through selection, crossover and mutation operations in the iterative process of the genetic algorithm; when optimizing the SVR model, the genetic algorithm is used to optimize the parameters of the SVR, including the regularization parameter C, the kernel function parameter γ and the error tolerance ε. The range of each parameter is set to (0, 100), (0, 1000), (0, 10). The optimization process is divided into the following steps Encoding: Encode the SVR parameters, regularization parameter C, kernel function parameter γ and error tolerance ε into chromosomes; Initialize the population: randomly generate a set of solutions as the initial population; Fitness calculation: Use cross-validation method to calculate the fitness of each individual, usually the prediction accuracy; Selection, crossover, and mutation: The chromosomes in the population evolve toward the optimal values ​​of C, γ, and ε through selection, crossover, and mutation, and generate new populations; Termination condition: Repeat the above process until the termination condition is met, the maximum number of iterations is reached, or the fitness reaches a certain threshold; After the optimal model is obtained by iteratively optimizing the support vector regression model parameters using a genetic algorithm, the optimal value of the objective function is predicted to obtain the best combination of pre-forging geometric parameters.

7. A high-temperature alloy disc pre-forging shape optimization system for implementing the high-temperature alloy disc pre-forging shape optimization method according to any one of claims 1 to 6, characterized in that: The high temperature alloy disc pre-forging shape optimization system comprises: Design module, used to simulate the disc-like parts without pre-forging process. According to the overall effect deformation of the disc-like parts after final forging, the reverse simulation technology is used to quickly design the initial geometric shape of the pre-forging parts, and the shape of the pre-forging die is designed according to the shape of the pre-forging parts; A screening module is used to combine the experimental design method with the finite element method to conduct screening experimental design to determine the characteristic geometric parameters of the pre-forging as key optimization parameters and determine the value range of the optimization parameters; The simulation module is used to determine the number of samples extracted by hypercube sampling and to automatically perform batch simulation on all random samples using the finite element simulation method; An extraction module is used to establish an objective function describing the forming quality and extract the required data from all sample simulation results as a data set for establishing a prediction model; The prediction module is used to establish a GA-SVR prediction model between input variables and objective functions through data sets, predict the optimal value of the objective function, and obtain the best combination of pre-forging geometric parameters.

8. A computer device, characterized in that: The computer device includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the high-temperature alloy disc pre-forging shape optimization method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the method for optimizing the shape of a high-temperature alloy disc pre-forging according to any one of claims 1 to 6.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the high-temperature alloy disc pre-forging shape optimization system as described in claim 7.

Images