Axial flow compressor blade profile optimization method based on main and auxiliary agent models and related device

Abstract

The invention discloses an axial flow compressor blade profile optimization method based on main and auxiliary agent models and a related device, and relates to the field of compressor blade design, and the method comprises the steps: generating a plurality of design parameter groups based on to-be-optimized design parameters and a parameter adjustable range of a to-be-optimized axial flow compressor blade by using a data sampling algorithm, outputting a blade profile performance prediction value corresponding to each design parameter group by using the main and auxiliary combination agent model, screening a plurality of alternative design parameter groups based on a comparison result of each blade profile performance prediction value and an initial performance reference, and performing computational fluid dynamics simulation; and determining the alternative design parameter group corresponding to the optimal simulation performance parameter as a blade profile optimization parameter group. According to the method, the invalid risk of the design parameter group is reduced by using the adjustable range of the parameters, the performance prediction precision is improved based on the flow field characteristic parameters by using the main and auxiliary combination agent model, finally, the blade profile optimization parameter group is determined based on the simulation result of the screened alternative design parameter group, and the prediction precision is further improved.

Description

Technical Field

[0001] This application relates to the field of compressor blade design technology, and in particular to an axial compressor blade profile optimization method and related apparatus based on a master-slave agent model. Background Technology

[0002] Axial compressors are critical equipment in the aerospace and energy fields, and their blade design directly determines their aerodynamic performance. Traditional design relies on engineers' experience to manually adjust parameters, which is time-consuming and prone to getting trapped in local optima. To address the high computational cost of existing technologies based on optimization algorithms and computational fluid dynamics (CFD) simulations, this paper proposes to construct a surrogate model to replace high-cost CFD analysis.

[0003] However, existing surrogate models only focus on the mapping between design variables and performance outputs, wasting a large amount of flow field data generated by CFD simulations. This leads to insufficient prediction accuracy in surrogate models trained with small sample sizes. Therefore, improving the prediction accuracy of surrogate models trained with small sample sizes for blade optimization parameters has become an urgent problem to be solved. Summary of the Invention

[0004] In view of the above problems, this application provides a method and related apparatus for axial compressor blade optimization based on a master-slave surrogate model, so as to improve the prediction accuracy of blade optimization parameters by the surrogate model trained with a small number of samples. The specific solution is as follows:

[0005] The first aspect of this application provides a method for optimizing the airfoil profile of an axial compressor based on a master-slave agent model, including:

[0006] The design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized are obtained. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized.

[0007] Based on the design parameters to be optimized and their adjustable range, multiple design parameter groups are generated using a data sampling algorithm.

[0008] Each of the design parameter groups is sequentially input into the primary and secondary combined surrogate model to obtain the predicted airfoil performance value corresponding to each of the design parameter groups. The training data of the primary and secondary combined surrogate model includes at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil.

[0009] The predicted performance values ​​of each blade profile are compared with the initial performance benchmark using a multi-objective optimization algorithm, and multiple alternative design parameter groups are selected from each design parameter group based on the comparison results.

[0010] Computational fluid dynamics simulations are performed based on each of the candidate design parameter groups to obtain the simulation performance parameters of each candidate design parameter group, and the candidate design parameter group corresponding to the optimal simulation performance parameters is determined as the airfoil optimization parameter group.

[0011] In one possible implementation, the training process of the master-slave combined agent model includes:

[0012] Obtain the sweep parameters of multiple sets of control points of the original airfoil. Each set of sweep parameters of the control points includes the sweep values ​​of each control point of the Bezier curve of the axial compressor blade to be optimized of the original airfoil.

[0013] Computational fluid dynamics simulations were performed based on the sweep parameters of each group of control points to obtain the stable operating range parameters and fluid characteristic parameters corresponding to the sweep parameters of each group of control points.

[0014] The dataset including the slant parameters of each group of control points with added labels is determined as the first training set, and the label content is the stable operating range parameter corresponding to the slant parameter of the control point; the dataset including each of the fluid feature parameters is determined as the second training set.

[0015] The initial primary and secondary agent model is trained based on the first training set and the second training set to obtain the primary and secondary combined agent model. The input of the primary and secondary combined agent model is a plurality of design parameter groups generated based on the design parameters to be optimized and the adjustable range of the parameters of the axial compressor blade to be optimized. The output is the blade performance prediction value corresponding to each design parameter group.

[0016] In one possible implementation, training the initial master-slave agent model based on the first training set and the second training set to obtain the master-slave combined agent model includes,

[0017] Initialize the initial master-slave agent model, which includes a master model, a slave model, a fusion layer, and a linear transformation layer;

[0018] Input the first training set into the main model to obtain the initial airfoil performance prediction results of the sweep parameters of each control point output by the main model;

[0019] The second training set is input into the sub-model to obtain the auxiliary airfoil performance prediction results of the sweep parameters of each control point output by the sub-model.

[0020] The initial and auxiliary airfoil performance prediction results are input into the fusion layer, so that the fusion layer sums the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights to obtain the airfoil performance prediction values ​​of the sweep parameters of each set of control points. Based on the airfoil performance prediction values ​​and the initial performance benchmark, the parameters of the main model and the sub-model are adjusted to obtain the main-sub combined surrogate model.

[0021] In one possible implementation, the predicted airfoil performance values ​​include at least a stable operating range and a peak efficiency prediction condition, and the initial performance baseline includes at least an initial stable operating range and an initial peak efficiency condition. The step of comparing each predicted airfoil performance value with the initial performance baseline using a multi-objective optimization algorithm, and selecting multiple candidate design parameter groups from each set of design parameters based on the comparison results, includes:

[0022] The multi-objective optimization algorithm is used to compare each of the predicted stable operating ranges with the initial stable operating range, and to compare each of the predicted peak efficiency conditions with the initial peak efficiency conditions. The design parameter group corresponding to the predicted airfoil performance values ​​where the predicted stable operating range is not less than the initial stable operating range and the predicted peak efficiency conditions are not less than the initial peak efficiency conditions is determined as the alternative design parameter group.

[0023] In one possible implementation, the fluid characteristic parameter is at least one or more of the following: high entropy region value, highest efficiency point, axial momentum of the axial compressor blade to be optimized, and aerodynamic load center location.

[0024] One possible implementation also includes:

[0025] In the absence of the alternative design parameter set, the design parameter set corresponding to each of the predicted performance values ​​of the blade profiles is used as supplementary training data, and the supplementary training data is used to optimize and train the main and auxiliary combined surrogate model.

[0026] Once the primary and secondary combined surrogate model has completed its optimization training, the operation step of sequentially inputting each of the design parameter groups into the primary and secondary combined surrogate model to obtain the predicted blade performance values ​​corresponding to each of the design parameter groups is executed.

[0027] A second aspect of this application provides an axial compressor blade optimization system based on a master-slave agent model, comprising:

[0028] The parameter acquisition module is used to obtain the design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized.

[0029] The parameter generation module is used to generate multiple design parameter groups based on the design parameters to be optimized and the adjustable range of the parameters using a data sampling algorithm;

[0030] The performance prediction module is used to input each of the design parameter groups into the main and auxiliary combined surrogate model in sequence to obtain the airfoil performance prediction value corresponding to each of the design parameter groups. The training data of the main and auxiliary combined surrogate model includes at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil.

[0031] The performance comparison module is used to compare the predicted performance values ​​of each of the blade profiles with the initial performance benchmark using a multi-objective optimization algorithm, and to select multiple alternative design parameter groups from each of the design parameter groups based on the comparison results.

[0032] The parameter output module is used to perform computational fluid dynamics simulation based on each of the candidate design parameter groups, obtain the simulation performance parameters of each of the candidate design parameter groups, and determine the candidate design parameter group corresponding to the optimal simulation performance parameters as the airfoil optimization parameter group.

[0033] In one possible implementation, the axial compressor blade optimization system further includes a model training module, which is configured during the training of the primary and secondary combined surrogate model as follows:

[0034] Obtain the sweep parameters of multiple sets of control points of the original airfoil. Each set of sweep parameters of the control points includes the sweep values ​​of each control point of the Bezier curve of the axial compressor blade to be optimized of the original airfoil.

[0035] Computational fluid dynamics simulations were performed based on the sweep parameters of each group of control points to obtain the stable operating range parameters and fluid characteristic parameters corresponding to the sweep parameters of each group of control points.

[0036] The dataset including the slant parameters of each group of control points with added labels is determined as the first training set, and the label content is the stable operating range parameter corresponding to the slant parameter of the control point; the dataset including each of the fluid feature parameters is determined as the second training set.

[0037] The initial primary and secondary agent model is trained based on the first training set and the second training set to obtain the primary and secondary combined agent model. The input of the primary and secondary combined agent model is a plurality of design parameter groups generated based on the design parameters to be optimized and the adjustable range of the parameters of the axial compressor blade to be optimized. The output is the blade performance prediction value corresponding to each design parameter group.

[0038] In one possible implementation, the model training module is configured to train the initial master-slave agent model based on the first training set and the second training set to obtain the master-slave combined agent model as follows:

[0039] Initialize the initial master-slave agent model, which includes a master model, a slave model, a fusion layer, and a linear transformation layer;

[0040] Input the first training set into the main model to obtain the initial airfoil performance prediction results of the sweep parameters of each control point output by the main model;

[0041] The second training set is input into the sub-model to obtain the auxiliary airfoil performance prediction results of the sweep parameters of each control point output by the sub-model.

[0042] The initial and auxiliary airfoil performance prediction results are input into the fusion layer, so that the fusion layer sums the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights to obtain the airfoil performance prediction values ​​of the sweep parameters of each set of control points. Based on the airfoil performance prediction values ​​and the initial performance benchmark, the parameters of the main model and the sub-model are adjusted to obtain the main-sub combined surrogate model.

[0043] In one possible implementation, the performance comparison module is configured as follows:

[0044] The multi-objective optimization algorithm is used to compare each stable operating prediction range with the initial stable operating range, and each peak efficiency prediction condition with the initial peak efficiency condition. The design parameter group corresponding to the airfoil performance prediction value where the stable operating prediction range is not less than the initial stable operating range and the peak efficiency prediction condition is not less than the initial peak efficiency condition is determined as the alternative design parameter group. The airfoil performance prediction value includes at least the stable operating prediction range and the peak efficiency prediction condition, and the initial performance benchmark includes at least the initial stable operating range and the initial peak efficiency condition.

[0045] In one possible implementation, the fluid characteristic parameter is at least one or more of the following: high entropy region value, highest efficiency point, axial momentum of the axial compressor blade to be optimized, and aerodynamic load center location.

[0046] In one possible implementation, the performance comparison module is further configured as follows:

[0047] In the absence of the alternative design parameter set, the design parameter set corresponding to each of the predicted performance values ​​of the blade profiles is used as supplementary training data, and the supplementary training data is used to optimize and train the main and auxiliary combined surrogate model.

[0048] Once the primary and secondary combined surrogate model has completed its optimization training, the performance prediction module is triggered to execute the operation step of sequentially inputting each of the design parameter groups into the primary and secondary combined surrogate model to obtain the blade performance prediction value corresponding to each of the design parameter groups.

[0049] A third aspect of this application provides a computer program product, including computer-readable instructions, which, when executed on an electronic device, cause the electronic device to implement the axial compressor blade optimization method based on a master-slave agent model as described in the first aspect or any implementation thereof.

[0050] A fourth aspect of this application provides an electronic device, including at least one processor and a memory connected to the processor, wherein:

[0051] The memory is used to store computer programs;

[0052] The processor is used to execute the computer program so that the electronic device can implement the axial compressor blade optimization method based on the master-slave agent model of the first aspect or any implementation of the first aspect described above.

[0053] The fifth aspect of this application provides a computer storage medium carrying one or more computer programs, which, when executed by an electronic device, enable the electronic device to implement the axial compressor blade optimization method based on a master-slave agent model as described in the first aspect or any implementation thereof.

[0054] By employing the aforementioned technical solution, this application provides a method and related apparatus for optimizing axial compressor blade profiles based on a master-slave surrogate model. This method determines the optimization object and reference standard by configuring candidate optimization design parameters affecting the aerodynamic performance of the axial compressor blade to be optimized, and initial performance benchmarks obtained through computational fluid dynamics simulation based on the original blade profile. Subsequently, by configuring a data sampling algorithm to generate multiple design parameter sets based on the design parameters to be optimized and their adjustable range, multiple parameter sets are generated. This reduces the risk of invalid parameters in the design parameter sets by utilizing the adjustable range, thereby improving the accuracy of subsequent prediction and optimization. Furthermore, by configuring the training data of the master-slave combined surrogate model to include at least the flow field characteristic parameters obtained from computational fluid dynamics simulation based on the original blade profile, the master-slave combined surrogate model can predict the performance of the design parameter sets based on the flow field characteristics captured during training. Subsequently, by configuring a multi-objective optimization algorithm, the predicted performance values ​​of each airfoil are compared with the initial performance benchmark. Based on the comparison results, multiple candidate design parameter groups are selected from each design parameter group, reducing the computational load of subsequent computational fluid dynamics simulations and improving the simulation accuracy of the obtained simulation performance parameters. Finally, by configuring computational fluid dynamics simulations based on each candidate design parameter group, the simulation performance parameters of each candidate design parameter group are obtained. The candidate design parameter group corresponding to the optimal simulation performance parameters is determined as the airfoil optimization parameter group, improving the prediction accuracy of the finally determined airfoil optimization parameter group. Attached Figure Description

[0055] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent when taken in conjunction with the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0056] Figure 1 A flowchart of an axial compressor blade profile optimization method based on a master-slave agent model provided for this application;

[0057] Figure 2 A schematic diagram of a Bézier curve provided for this application;

[0058] Figure 3 A schematic diagram of an initial master-slave agent model provided for this application;

[0059] Figure 4 A schematic diagram of a high-entropy region of a blade provided in this application;

[0060] Figure 5 A flow-pressure ratio characteristic curve is provided for this application;

[0061] Figure 6This application provides a flow-isentropic efficiency characteristic curve;

[0062] Figure 7 A block diagram of an axial compressor blade profile optimization system provided in this application;

[0063] Figure 8 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0064] The embodiments of this application are described below with reference to the accompanying drawings. The terminology used in the implementation section of this application is for explaining specific embodiments only and is not intended to limit the scope of this application.

[0065] The embodiments of this application will now be described with reference to the accompanying drawings. Those skilled in the art will recognize that, with technological advancements and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are equally applicable to similar technical problems.

[0066] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0067] The first aspect of this application provides a method for optimizing the airfoil profile of an axial compressor based on a master-slave agent model, such as... Figure 1 As shown, the axial compressor blade optimization method based on the master-slave agent model includes:

[0068] S101. Obtain the design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized.

[0069] It should be noted that, in practical applications, the aforementioned adjustable geometric parameters of the blade refer to the sweep values ​​at various control points on the Bézier curve of the axial compressor blade to be optimized. The sweep value indicates the radial tilt angle of the blade; a positive sweep value indicates the blade is tilted axially towards the gas outlet, while a negative sweep value indicates the blade is tilted in the opposite direction axially towards the gas outlet. Since the shape of the axial compressor blade to be optimized is set to a regular shape (e.g., rectangular) or an irregular shape (e.g., arc or curved surface) based on its design requirements, the sweep value at a single point is insufficient to characterize the aerodynamic performance of the entire blade profile when the blade shape is irregular. Therefore, by equally dividing the Bézier curve of the axial compressor blade to be optimized into multiple control points and extracting the sweep value of each control point, the accuracy of the subsequent prediction results is improved.

[0070] In one possible implementation, there are multiple ways to obtain the Bezier curve of the axial compressor blade to be optimized and the control points thereon. Here, one example is provided:

[0071] like Figure 2 The diagram shows a schematic of the Bézier curve for an axial compressor blade to be optimized. Obtaining the aforementioned Bézier curve can include steps A1 to A3 as described below.

[0072] Step A1: Divide the axial compressor blades to be optimized into 10 sections radially. Then trigger step A2.

[0073] In one possible implementation, the 10 sections in step A1 above are as follows: Figure 2 The sections specified by numbers 1 to 10. Section 1 is the section at the blade root, and section 10 is the section at the blade tip. The direction from the centroid of section 1 to the centroid of section 10 is the radial direction in step A1 above.

[0074] In one possible implementation, step A1 above can be achieved using the "radial plane cutting" function in a modeling tool. These modeling tools include, but are not limited to: Computer-Aided Design (CAD) tools, Unigraphics (UG) tools, SolidWorks, etc.

[0075] Step A2: Locate the geometric center (or centroid / center of mass) of each section in Step A1, and determine the curves fitted to each geometric center (or centroid / center of mass) as the Bezier curves of the axial compressor blade to be optimized. Then trigger Step A3.

[0076] In one possible implementation, the aforementioned Bézier curve is as follows: Figure 2 The broken line segment running through each cross section is shown.

[0077] Step A3: Divide the Bézier curve into equal intervals and determine each division point as a control point of the Bézier curve.

[0078] In one possible implementation, the control points of the aforementioned Bézier curve are as follows: Figure 2 The inflection points on the Bézier curve are shown.

[0079] It should be noted that, in practical application scenarios, this application obtains an initial performance standard by configuring and performing computational fluid dynamics simulation based on the original airfoil of the axial compressor blade to be optimized. This enables the acquisition of relatively realistic and accurate performance standard parameters. Subsequently, the prediction results are verified using this initial performance standard, avoiding the risk that the output prediction results are lower than the original airfoil performance of the axial compressor blade to be optimized, and improving the accuracy of the final predicted airfoil optimization parameter set.

[0080] In one possible implementation, the initial performance benchmark may include entropy efficiency parameters (e.g., the increase in stable operating range and peak efficiency) corresponding to each control point during the computational fluid dynamics simulation of the axial compressor blade to be optimized. The increase in stable operating range I can be calculated using the following formula: . It is the stable operating range of the optimized blade profile. This is the stable operating range of the original blade shape.

[0081] S102. Using a data sampling algorithm, generate multiple design parameter groups based on the design parameters to be optimized and the adjustable range of the parameters.

[0082] It should be noted that in practical applications, the above-mentioned data sampling algorithms can be of various types, including but not limited to: Latin Hypercube Sampling (LHS), Monte Carlo Sampling (MC), and Uniform Sampling (US). This application does not impose excessive restrictions or elaborate on the types of the above-mentioned data sampling algorithms.

[0083] It should be noted that in practical applications, the adjustable range of the above parameters refers to the range of adjustment of the adjustable geometric parameters of the blade without causing the blade performance to fall below the initial performance baseline. The adjustable range of these parameters can be obtained through calibration experiments on the axial compressor blade to be optimized. For example, the adjustable range of the above parameters can be -5° to +5°. It should also be noted that for axial compressor blades to be optimized with the same or similar shapes but different dimensions, the same adjustable range of parameters can be used, or they can be calibrated independently.

[0084] It should be noted that, in practical applications, the aforementioned design parameter set may include random sweep values ​​of each control point in the design parameters to be optimized. All of these random sweep values ​​are within the adjustable range of the parameters.

[0085] It should be noted that this application generates multiple design parameter sets based on the design parameters to be optimized and the adjustable range of the parameters by configuring and utilizing a data sampling algorithm. The adjustable range of the parameters is used to constrain the parameters in the design parameter sets, thereby avoiding the risk of obtaining design parameter sets containing parameters that would lead to further deterioration of blade performance, and thus improving the accuracy of the subsequently predicted blade shape optimization parameter sets.

[0086] S103. Input each design parameter group into the main and auxiliary combined surrogate model in sequence to obtain the airfoil performance prediction value corresponding to each design parameter group. The training data of the main and auxiliary combined surrogate model shall include at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil.

[0087] It should be noted that in practical applications, the aforementioned flow field characteristic parameters are core indicators that quantitatively describe the airflow state, flow pattern, and energy conversion process passing through the blade under its influence. These parameters characterize the blade's aerodynamic performance. Because existing surrogate models only focus on the mapping between design variables and performance outputs, they cannot capture the correlation between flow field characteristics and blade aerodynamic performance. This results in an inability to fully capture the correlation characteristics of image blade aerodynamic performance in small-sample training scenarios, reducing the prediction accuracy of the surrogate model for blade optimization parameters. Therefore, this application configures a master-slave combined surrogate model trained with training data including at least the flow field characteristic parameters obtained from computational fluid dynamics simulations based on the original airfoil. Based on the input design parameter groups, it outputs the corresponding airfoil performance prediction values ​​for each design parameter group. This allows the master-slave combined surrogate model to predict the performance of the design parameter groups based on the correlation between the flow field characteristics captured during training and the blade aerodynamic performance, improving the accuracy of the output airfoil performance prediction values.

[0088] S104. Use a multi-objective optimization algorithm to compare the predicted performance values ​​of each airfoil with the initial performance benchmark, and select multiple alternative design parameter groups from each design parameter group based on the comparison results.

[0089] It should be noted that in practical applications, any prediction model is susceptible to a decrease in accuracy due to overgeneralization. Therefore, this application utilizes a multi-objective optimization algorithm to compare the predicted performance values ​​of each airfoil with the initial performance benchmark. Based on the comparison results, multiple candidate design parameter groups are selected from each design parameter group to eliminate abnormal design parameter groups, thereby further improving the accuracy of the final determined airfoil optimization parameter group.

[0090] It should be noted that in practical applications, the aforementioned Multi-Objective Optimization Algorithm (MOA) is used to find the optimal solution when considering multiple conflicting objectives. There are various types of MOA, including but not limited to: Non-dominated Sorting Genetic Algorithm II (NSGA-II), Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA / D), and Strength Pareto Evolutionary Algorithm 2 (SPEA2). Since both the predicted airfoil performance values ​​and the initial performance benchmark have various types of data, relying solely on a single data set for selection risks failing to accurately determine the optimal airfoil optimization parameter set. Therefore, this application utilizes a multi-objective optimization algorithm to compare the predicted airfoil performance values ​​with the initial performance benchmark, and based on the comparison results, selects multiple candidate design parameter sets from each design parameter set, thereby improving the accuracy of the ultimately determined airfoil optimization parameter set.

[0091] S105. Perform computational fluid dynamics simulation based on each alternative design parameter group to obtain the simulation performance parameters of each alternative design parameter group, and determine the alternative design parameter group corresponding to the optimal simulation performance parameters as the airfoil optimization parameter group.

[0092] It should be noted that this application further verifies the performance of each alternative design parameter group by configuring computational fluid dynamics simulation based on each alternative design parameter group, thereby ensuring that the final determined blade optimization parameter group is the optimal set of optimization parameters that makes the performance of the axial compressor blade to be optimized the best. As a result, compared with the prior art, this application improves the prediction accuracy of the surrogate model trained with small sample size for blade optimization parameters.

[0093] This application achieves the determination of the optimization object and reference standard by configuring and obtaining candidate optimization design parameters affecting the aerodynamic performance of the axial compressor blade to be optimized, and initial performance benchmarks obtained from computational fluid dynamics simulation based on the original blade profile. Subsequently, by configuring and utilizing a data sampling algorithm based on the design parameters to be optimized and their adjustable range, multiple design parameter groups are generated. This reduces the risk of invalid parameters in the design parameter groups by utilizing the adjustable range, thereby improving the accuracy of subsequent prediction and optimization. Furthermore, by configuring the training data of the master-slave combined surrogate model to include at least the flow field characteristic parameters obtained from computational fluid dynamics simulation based on the original blade profile, the master-slave combined surrogate model can predict the performance of the design parameter groups based on the flow field characteristics captured during training. Subsequently, by configuring and utilizing a multi-objective optimization algorithm, the predicted performance values ​​of each blade profile are compared with the initial performance benchmark, and multiple candidate design parameter groups are selected from each design parameter group based on the comparison results. This reduces the computational load of subsequent computational fluid dynamics simulations and improves the simulation accuracy of the subsequently obtained simulation performance parameters. Finally, by configuring computational fluid dynamics simulation based on each alternative design parameter group, the simulation performance parameters of each alternative design parameter group are obtained, and the alternative design parameter group corresponding to the optimal simulation performance parameters is determined as the airfoil optimization parameter group, thereby improving the prediction accuracy of the finally determined airfoil optimization parameter group.

[0094] In one possible implementation, the training process of the aforementioned master-slave combined agent model includes:

[0095] Obtain the sweep parameters of multiple control points of the original airfoil. The sweep parameters of a set of control points include the sweep values ​​of each control point of the Bezier curve of the axial compressor blade to be optimized of the original airfoil.

[0096] Computational fluid dynamics simulations were performed based on the sweep parameters of each set of control points to obtain the stable operating range parameters and fluid characteristic parameters corresponding to the sweep parameters of each set of control points.

[0097] The dataset including the sweep parameters of each group of control points with added labels is determined as the first training set, and the labels are the stable working range parameters corresponding to the sweep parameters of the control points; the dataset including each fluid characteristic parameter is determined as the second training set.

[0098] The initial primary and secondary agent models are trained based on the first and second training sets to obtain the primary and secondary combined agent model. The input of the primary and secondary combined agent model is multiple design parameter sets generated based on the design parameters to be optimized and the adjustable range of the parameters of the axial compressor blade to be optimized. The output is the blade performance prediction value corresponding to each design parameter set.

[0099] It should be noted that, in practical application scenarios, the aforementioned stable operating range parameter M... SWThe near-surge mass flow rate M at each control point during the simulation can be used as a basis. NC Near stall point mass flow rate M NS The difference is shown in the formula: M SW =M NC -M NS .

[0100] In one possible implementation, the initial master-slave agent model is trained based on a first training set and a second training set to obtain a master-slave combined agent model, including:

[0101] Initialize the initial master and slave agent model, which includes the master model, slave model, fusion layer and linear transformation layer;

[0102] Input the first training set into the main model to obtain the initial airfoil performance prediction results of the sweep parameters of each control point output by the main model;

[0103] Input the second training set into the sub-model to obtain the auxiliary airfoil performance prediction results of the sweep parameters of each control point output by the sub-model;

[0104] The initial and auxiliary airfoil performance prediction results are input into the fusion layer. The fusion layer sums the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights to obtain the airfoil performance prediction values ​​of the sweep parameters of each set of control points. Based on the airfoil performance prediction values ​​and the initial performance benchmark, the parameters of the main model and the sub-model are adjusted to obtain the main and sub-model combined surrogate model.

[0105] It should be noted that in practical applications, the initial master and slave agent models described above can be built based on a multilayer perceptron (MLP) architecture. Since the amount of training data in the first training set is limited in small-sample scenarios, and the limited training data can lead to overfitting risks in ordinary MLP architectures, the master model can be configured as a regularized MLP architecture. This leverages regularization to improve the master model's generalization ability in small-sample scenarios, avoiding the risk of the master model missing features due to overfitting and improving its output accuracy. Furthermore, since the slave model has fewer input dimensions, to improve the overall prediction efficiency of the master-slave combined agent model and avoid the risk of decreased prediction accuracy due to increased fit caused by the complexity of the slave model structure, the slave model can be configured as a lightweight MLP architecture with lightweight feature enhancement channels. This ensures the output accuracy of the slave model while avoiding the risk of increased overall fit in the master-slave combined agent model.

[0106] In one possible implementation, the mathematical representation of the above linear transformation layer can be shown in the following formula: ,in, It is the summation weight of the initial or auxiliary airfoil performance prediction results. It is a weight matrix. It is a bias vector. Each (The sum of the initial and auxiliary airfoil performance predictions corresponding to the sweep parameters of a set of control points). During the process of adjusting the summation weights by calling the linear transformation layer, the above is first solved based on the initial and auxiliary airfoil performance predictions from the input fusion layer. And call the above linear transformation layer to solve Then the Determined as above During the solution process The summation weights of main, and... The summation weights of aux are set to Then, the sum of the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of each control point is calculated. Since the importance of the output results of the main model and the sub-model differs under different input conditions, this application improves the output accuracy of the blade performance prediction value by configuring the above-mentioned linear transformation layer to dynamically adjust the summation weights.

[0107] In one possible implementation, the architecture diagram of the initial master-slave agent model described above is as follows: Figure 3 As shown, the main model inputs the initial airfoil performance prediction results into the fusion layer, and the sub-model inputs the auxiliary airfoil performance prediction results into the fusion layer. The fusion layer sums the initial airfoil performance prediction results and the auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights during the summation process, and outputs the airfoil performance prediction values ​​of the sweep parameters of each set of control points.

[0108] In one possible implementation, the predicted airfoil performance values ​​include at least the predicted stable operating range and the predicted peak efficiency condition, and the initial performance baseline includes at least the initial stable operating range and the initial peak efficiency condition. A multi-objective optimization algorithm is used to compare the predicted airfoil performance values ​​with the initial performance baseline, and based on the comparison results, multiple candidate design parameter sets are selected from each set of design parameters, including:

[0109] Using a multi-objective optimization algorithm, the predicted range of each stable operating condition is compared with the initial stable operating range, and the predicted peak efficiency condition is compared with the initial peak efficiency condition. The design parameter group corresponding to the predicted airfoil performance values ​​that are not less than the initial stable operating range and whose predicted peak efficiency condition is not less than the initial peak efficiency condition is determined as the alternative design parameter group.

[0110] It should be noted that, in practical applications, the peak efficiency in both the predicted peak efficiency condition and the initial peak efficiency condition can be calculated using the following formula: Where m is the mass flow rate; η PE This is the peak efficiency value under the corresponding operating conditions; π t It is the peak efficiency point pressure ratio; T t，in It is the total imported temperature; P in It is the total inlet pressure; c p is the static pressure coefficient; k is the specific heat ratio.

[0111] In one possible implementation, the fluid characteristic parameters are at least one or more of the following: high-entropy region values, axial momentum of the axial compressor blades to be optimized at the highest efficiency point, and location of the aerodynamic load center.

[0112] It should be noted that in practical applications, the aforementioned high-entropy region values ​​can be based on the average width of the tail of the high-entropy region at different blade heights (e.g., 25%, 50%, and 75% blade height) in the flow field entropy diagram of the simulation results, where the entropy value exceeds a preset threshold. A schematic diagram of the high-entropy region of the blade can be shown below. Figure 4 As shown, the high-entropy region characterizes the degree of irreversible energy dissipation and the smoothness and disorder of the airflow behind the blade trailing edge after the airflow flows along the blade surface from the leading edge to the trailing edge. Figure 4 The values ​​for the medium-to-high entropy region are the default. Figure 4 The high-entropy region is defined as a region with an entropy value greater than a preset threshold, and the width value corresponding to the average width of the tail of this high-entropy region. The axial momentum of the axial compressor blade at the highest efficiency point can be calculated using the following formula: (The formula is missing from the provided text.) , of which M Z,in It is the dominant axial momentum; V z ρ is the axial velocity; m is the fluid density; and ρ is the reference flow rate. The aerodynamic load center position is obtained as follows: starting from the leading edge of the airfoil, the pressure difference between the suction and pressure surfaces is integrated. When this integral value reaches half of the total pressure difference across the entire airfoil cross-section, the corresponding axial coordinate is the aerodynamic load center position. In other words, at a certain location on the airfoil where the aerodynamic load center position is located, the integral of the pressure difference from the leading edge to that location accounts for 50% of the total pressure difference across the entire airfoil cross-section.

[0113] One possible implementation also includes:

[0114] In the absence of alternative design parameter sets, the design parameter sets corresponding to the predicted performance values ​​of each airfoil are used as supplementary training data, and the main and auxiliary combined surrogate model is optimized and trained using the supplementary training data.

[0115] Once the primary and secondary combined surrogate model has completed its optimization training, the operation steps are performed to sequentially input each design parameter group into the primary and secondary combined surrogate model to obtain the predicted airfoil performance values ​​corresponding to each design parameter group.

[0116] It should be noted that in practical application scenarios, this application configures the design parameter group corresponding to the performance prediction value of each blade type as supplementary training data when there is no alternative design parameter group. The supplementary training data is used to optimize and train the main and auxiliary combined surrogate model, thereby realizing the dynamic optimization of the main and auxiliary combined surrogate model in small sample scenarios, and thus improving the output accuracy of the main and auxiliary combined surrogate model.

[0117] The researchers conducted numerical simulations on a transonic axial compressor with 16 channels to verify the optimization effect of the axial compressor blade profile optimization method based on the master-slave proxy model provided in this application. To save computational costs, a single channel was selected for numerical simulation. AutoGrid5 in NUMECA software was used to mesh the blade channels. The meshed compressor was imported into CFX preprocessing, with total temperature and total pressure as inlet conditions, static pressure outlet as outlet boundary conditions, and SST (Shear Stress Transfer) as the turbulence model. Then, CFD-post was used for post-processing of the numerical calculation results to obtain the optimization results. Comparative analysis of the experimental results yielded the following results: Figure 5 The flow-pressure ratio characteristic curves of the optimized airfoil (optimized airfoil) and the original airfoil (prototype) are shown, as well as... Figure 6 The flow-isentropic efficiency characteristic curves of the optimized airfoil (optimized airfoil) and the original airfoil (prototype) are shown in the two figures. The margin increase value in the two figures represents the improvement in stall margin of the optimized airfoil compared to the original airfoil. The two broken lines in the two figures represent the overall stable operating range of the original airfoil and the optimized airfoil, respectively. The larger this range, the better. However, while the range of the x-coordinate (flow coefficient) is expanded, the total pressure ratio and isentropic efficiency of the y-coordinate should not decrease significantly. The research results show that, based on a small sample library, the master-slave combined surrogate model in this invention improves the prediction accuracy by 61.2% compared to the traditional surrogate model. The optimized airfoil has a 2.91° forward sweep at 95% blade height and a 2.63° backward sweep at 50% blade height. The peak efficiency of the optimized airfoil is improved by 0.5%, and the stable operating range is increased by 17.7%. This verifies the applicability of the master-slave combined surrogate model in compressor performance optimization design. The second aspect of this application provides an optimization system for axial compressor airfoils, such as... Figure 7 As shown, the optimization system for the axial compressor blade profile includes:

[0118] The parameter acquisition module 701 is used to obtain the design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized.

[0119] The parameter generation module 702 is used to generate multiple design parameter groups based on the design parameters to be optimized and the adjustable range of the parameters using a data sampling algorithm;

[0120] The performance prediction module 703 is used to input each design parameter group into the main and auxiliary combined surrogate model in sequence to obtain the airfoil performance prediction value corresponding to each design parameter group. The training data of the main and auxiliary combined surrogate model includes at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil.

[0121] The performance comparison module 704 is used to compare the predicted performance values ​​of each airfoil with the initial performance benchmark using a multi-objective optimization algorithm, and to select multiple alternative design parameter groups from each design parameter group based on the comparison results.

[0122] The parameter output module 705 is used to perform computational fluid dynamics simulation based on each alternative design parameter group, obtain the simulation performance parameters of each alternative design parameter group, and determine the alternative design parameter group corresponding to the optimal simulation performance parameters as the airfoil optimization parameter group.

[0123] In one possible implementation, the axial compressor blade optimization system provided in the second aspect of this application further includes a model training module, which is configured during the training of the primary and secondary combined surrogate model as follows:

[0124] Obtain the sweep parameters of multiple control points of the original airfoil. The sweep parameters of a set of control points include the sweep values ​​of each control point of the Bezier curve of the axial compressor blade to be optimized of the original airfoil.

[0125] Computational fluid dynamics simulations were performed based on the sweep parameters of each set of control points to obtain the stable operating range parameters and fluid characteristic parameters corresponding to the sweep parameters of each set of control points.

[0126] The dataset including the sweep parameters of each group of control points with added labels is determined as the first training set, and the labels are the stable working range parameters corresponding to the sweep parameters of the control points; the dataset including each fluid characteristic parameter is determined as the second training set.

[0127] The initial primary and secondary agent models are trained based on the first and second training sets to obtain the primary and secondary combined agent model. The input of the primary and secondary combined agent model is multiple design parameter sets generated based on the design parameters to be optimized and the adjustable range of the parameters of the axial compressor blade to be optimized. The output is the blade performance prediction value corresponding to each design parameter set.

[0128] In one possible implementation, the above-mentioned model training module is configured to train the initial master-slave agent model based on the first and second training sets to obtain the master-slave combined agent model as follows:

[0129] Initialize the initial master and slave agent model, which includes the master model, slave model, fusion layer and linear transformation layer;

[0130] Input the first training set into the main model to obtain the initial airfoil performance prediction results of the sweep parameters of each control point output by the main model;

[0131] Input the second training set into the sub-model to obtain the auxiliary airfoil performance prediction results of the sweep parameters of each control point output by the sub-model;

[0132] The initial and auxiliary airfoil performance prediction results are input into the fusion layer. The fusion layer sums the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights to obtain the airfoil performance prediction values ​​of the sweep parameters of each set of control points. Based on the airfoil performance prediction values ​​and the initial performance benchmark, the parameters of the main model and the sub-model are adjusted to obtain the main and sub-model combined surrogate model.

[0133] In one possible implementation, the performance comparison module 704 described above is configured as follows:

[0134] A multi-objective optimization algorithm is used to compare each stable operating prediction range with the initial stable operating range, and each peak efficiency prediction condition with the initial peak efficiency condition. The design parameter group corresponding to the airfoil performance prediction values ​​with a stable operating prediction range not less than the initial stable operating range and a peak efficiency prediction condition not less than the initial peak efficiency condition is determined as the candidate design parameter group. The airfoil performance prediction values ​​include at least the stable operating prediction range and the peak efficiency prediction condition, and the initial performance benchmark includes at least the initial stable operating range and the initial peak efficiency condition.

[0135] In one possible implementation, the fluid characteristic parameters are at least one or more of the following: high-entropy region values, axial momentum of the axial compressor blades to be optimized at the highest efficiency point, and location of the aerodynamic load center.

[0136] In one possible implementation, the performance comparison module 704 is further configured as follows:

[0137] In the absence of alternative design parameter sets, the design parameter sets corresponding to the predicted performance values ​​of each airfoil are used as supplementary training data, and the main and auxiliary combined surrogate model is optimized and trained using the supplementary training data.

[0138] Once the primary and secondary combined surrogate model has completed its optimization training, the performance prediction module is triggered to execute the operation steps of sequentially inputting each design parameter group into the primary and secondary combined surrogate model to obtain the corresponding blade performance prediction values ​​for each design parameter group.

[0139] A third aspect of this application provides a computer program product, including computer-readable instructions, which, when executed on an electronic device, cause the electronic device to implement the axial compressor blade optimization method based on a master-slave agent model as described in the first aspect or any implementation thereof.

[0140] A fourth aspect of this application provides an electronic device, including at least one processor and a memory connected to the processor, wherein:

[0141] Memory is used to store computer programs;

[0142] The processor is used to execute computer programs to enable electronic devices to implement the axial compressor blade optimization method based on the master-slave agent model, which is described in the first aspect or any implementation of the first aspect above.

[0143] The fifth aspect of this application provides a computer storage medium carrying one or more computer programs, which, when executed by an electronic device, enable the electronic device to implement the axial compressor blade optimization method based on a master-slave agent model as described in the first aspect or any implementation thereof.

[0144] The structural schematic diagram of the electronic device provided in the fourth aspect of this application is as follows: Figure 8 As shown. The electronic devices in the embodiments of this application may include, but are not limited to, fixed terminals such as mobile phones, laptops, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Figure 8 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0145] like Figure 8As shown, the electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage device 808 into a random access memory (RAM) 803. When the electronic device is powered on, the RAM 803 also stores various programs and data required for the operation of the electronic device. The processing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0146] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 808 including, for example, memory cards, hard drives, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0147] It should also be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the device embodiment drawings provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0148] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, training equipment, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0149] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0150] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

Claims

1. A method for optimizing the airfoil profile of an axial compressor based on a master-slave agent model, characterized in that, include: The design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized are obtained. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized. Based on the design parameters to be optimized and their adjustable range, multiple design parameter groups are generated using a data sampling algorithm. Each of the design parameter groups is sequentially input into the primary and secondary combined surrogate model to obtain the predicted airfoil performance value corresponding to each of the design parameter groups. The training data of the primary and secondary combined surrogate model includes at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil. The predicted performance values ​​of each blade profile are compared with the initial performance benchmark using a multi-objective optimization algorithm, and multiple alternative design parameter groups are selected from each design parameter group based on the comparison results. Computational fluid dynamics simulations are performed based on each of the candidate design parameter groups to obtain the simulation performance parameters of each candidate design parameter group, and the candidate design parameter group corresponding to the optimal simulation performance parameters is determined as the airfoil optimization parameter group.

2. The method for optimizing axial compressor blade profiles based on a master-slave agent model according to claim 1, characterized in that, The training process of the master-slave combined agent model includes: Obtain the sweep parameters of multiple sets of control points of the original airfoil. Each set of sweep parameters of the control points includes the sweep values ​​of each control point of the Bezier curve of the axial compressor blade to be optimized of the original airfoil. Computational fluid dynamics simulations were performed based on the sweep parameters of each group of control points to obtain the stable operating range parameters and fluid characteristic parameters corresponding to the sweep parameters of each group of control points. The dataset including the slant parameters of each group of control points with added labels is determined as the first training set, and the label content is the stable operating range parameter corresponding to the slant parameter of the control point; the dataset including each of the fluid feature parameters is determined as the second training set. The initial primary and secondary agent model is trained based on the first training set and the second training set to obtain the primary and secondary combined agent model. The input of the primary and secondary combined agent model is a plurality of design parameter groups generated based on the design parameters to be optimized and the adjustable range of the parameters of the axial compressor blade to be optimized. The output is the blade performance prediction value corresponding to each design parameter group.

3. The axial compressor blade optimization method based on the master-slave agent model according to claim 2, characterized in that, The step of training the initial master-slave agent model based on the first training set and the second training set to obtain the master-slave combined agent model includes, Initialize the initial master-slave agent model, which includes a master model, a slave model, a fusion layer, and a linear transformation layer; Input the first training set into the main model to obtain the initial airfoil performance prediction results of the sweep parameters of each control point output by the main model; The second training set is input into the sub-model to obtain the auxiliary airfoil performance prediction results of the sweep parameters of each control point output by the sub-model. The initial and auxiliary airfoil performance prediction results are input into the fusion layer, so that the fusion layer sums the initial and auxiliary airfoil performance prediction results corresponding to the sweep parameters of a set of control points, and calls the linear transformation layer to adjust the summation weights to obtain the airfoil performance prediction values ​​of the sweep parameters of each set of control points. Based on the airfoil performance prediction values ​​and the initial performance benchmark, the parameters of the main model and the sub-model are adjusted to obtain the main-sub combined surrogate model.

4. The method for optimizing axial compressor blade profiles based on a master-slave agent model according to claim 1, characterized in that, The predicted airfoil performance values ​​include at least a stable operating range and a peak efficiency prediction condition, and the initial performance baseline includes at least an initial stable operating range and an initial peak efficiency condition. The process involves comparing each predicted airfoil performance value with the initial performance baseline using a multi-objective optimization algorithm, and selecting multiple candidate design parameter groups from each set of design parameters based on the comparison results, including: The multi-objective optimization algorithm is used to compare each of the predicted stable operating ranges with the initial stable operating range, and to compare each of the predicted peak efficiency conditions with the initial peak efficiency conditions. The design parameter group corresponding to the predicted airfoil performance values ​​where the predicted stable operating range is not less than the initial stable operating range and the predicted peak efficiency conditions are not less than the initial peak efficiency conditions is determined as the alternative design parameter group.

5. The method for optimizing axial compressor blade profiles based on a master-slave agent model according to claim 2, characterized in that, The fluid characteristic parameters are at least one or more of the following: high entropy region value, highest efficiency point, axial momentum of the axial compressor blade to be optimized, and aerodynamic load center location.

6. The axial compressor blade optimization method based on the master-slave agent model according to claim 4, characterized in that, Also includes: In the absence of the alternative design parameter set, the design parameter set corresponding to each of the predicted performance values ​​of the blade profiles is used as supplementary training data, and the supplementary training data is used to optimize and train the main and auxiliary combined surrogate model. Once the primary and secondary combined surrogate model has completed its optimization training, the operation step of sequentially inputting each of the design parameter groups into the primary and secondary combined surrogate model to obtain the predicted blade performance values ​​corresponding to each of the design parameter groups is executed.

7. An axial compressor blade profile optimization system based on a master-slave agent model, characterized in that, include: The parameter acquisition module is used to obtain the design parameters to be optimized and the initial performance benchmark of the axial compressor blade to be optimized. The design parameters to be optimized are the adjustable geometric parameters of the blade that affect the aerodynamic performance of the blade profile. The initial performance benchmark is the result parameter obtained by computational fluid dynamics simulation based on the original blade profile of the axial compressor blade to be optimized. The parameter generation module is used to generate multiple design parameter groups based on the design parameters to be optimized and the adjustable range of the parameters using a data sampling algorithm; The performance prediction module is used to input each of the design parameter groups into the main and auxiliary combined surrogate model in sequence to obtain the airfoil performance prediction value corresponding to each of the design parameter groups. The training data of the main and auxiliary combined surrogate model includes at least the flow field characteristic parameters obtained by computational fluid dynamics simulation based on the original airfoil. The performance comparison module is used to compare the predicted performance values ​​of each of the blade profiles with the initial performance benchmark using a multi-objective optimization algorithm, and to select multiple alternative design parameter groups from each of the design parameter groups based on the comparison results. The parameter output module is used to perform computational fluid dynamics simulation based on each of the candidate design parameter groups, obtain the simulation performance parameters of each of the candidate design parameter groups, and determine the candidate design parameter group corresponding to the optimal simulation performance parameters as the airfoil optimization parameter group.

8. A computer program product, characterized in that, Includes computer-readable instructions that, when executed on an electronic device, cause the electronic device to implement the axial compressor blade optimization method based on a master-slave agent model as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, It includes at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the axial compressor blade optimization method based on the master-slave agent model as described in any one of claims 1 to 6.

10. A computer storage medium, characterized in that, The storage medium carries one or more computer programs that, when executed by an electronic device, enable the electronic device to implement the axial compressor blade optimization method based on the master-slave agent model as described in any one of claims 1 to 6.

Images