Methodology for developing an airworthiness compliance verification test matrix

Abstract

The purpose of this invention is to provide a method for formulating an airworthiness compliance verification test matrix. By employing a joint machine learning model and an iterative test dataset, the prediction accuracy of the machine learning model on the test set is used as an indicator. The test dataset is continuously updated until the prediction accuracy meets the indicator, and the set of training operating points is output as the airworthiness compliance verification test matrix for aero-engines. This ensures that the airworthiness compliance verification test matrix can cover the vibration characteristics of the blades throughout the entire flight envelope. This method for formulating an airworthiness compliance verification test matrix can reduce the number of test operating conditions while obtaining the dynamic stress response of the engine blades throughout the declared flight envelope.

Description

Technical Field

[0001] This invention relates to the field of gas turbine engine technology, and in particular to a method for developing an airworthiness compliance verification test matrix. Background Technology

[0002] The requirements of CCAR 33.83 airworthiness regulation stipulate that civil aircraft engines must undergo vibration stress testing to determine whether the vibration characteristics of components that may be excited by mechanical or aerodynamic forces are acceptable throughout the declared flight envelope. Furthermore, after allowing appropriate tolerances for material performance variations under various operating conditions, the sum of the vibration stress and the appropriate steady-state stress must be less than the relevant material's endurance limit.

[0003] Therefore, vibration stress testing of aero-engines needs to cover the entire declared flight envelope, involving numerous test conditions, long test cycles, high risks, and high costs. To ensure the completeness and accuracy of test results, stringent requirements are imposed on test design and result analysis methods. However, testing all operating points within the flight envelope would drastically extend the test cycle, increase test costs, and amplify the risks associated with engine development.

[0004] Therefore, it is necessary to design a reasonable test matrix to reduce the number of test conditions while obtaining the dynamic stress response of the engine blades within the entire declared flight envelope, thereby verifying the airworthiness compliance of the engine blades with respect to CCAR33.83. Summary of the Invention

[0005] The purpose of this invention is to provide a method for formulating an airworthiness compliance verification test matrix, which can reduce the number of test conditions while obtaining the dynamic stress response of engine blades within the entire declared flight envelope.

[0006] The method for developing an airworthiness compliance verification test matrix to achieve the aforementioned objectives includes the following steps:

[0007] a. Obtain the flight envelope declared by the engine;

[0008] b. Normalize the parameters according to the range set for all operating condition parameters within the flight envelope;

[0009] c. Obtain the test condition point;

[0010] d. Obtain the initial test condition point and the set of test condition points, and put the initial test condition point into the set of test condition points;

[0011] e. Conduct tests based on the set of test operating points and establish a test dataset;

[0012] f. Train the turbine blade dynamic stress prediction model;

[0013] g. Extract the test condition point set into a training condition point set and a test condition point set according to any ratio, and use the test dataset of the test program graph corresponding to the training condition point set as the training set, and use the test dataset of the test program graph corresponding to the test condition point set as the test set.

[0014] h. Determine whether the prediction accuracy of the turbine blade dynamic stress prediction model on the test set meets the requirements. If it does, output the blade dynamic stress prediction model and the training operating point set. If it does not meet the requirements, proceed to step i.

[0015] i. Extract the worst-case scenario point from the set of test scenario points where the model prediction accuracy is the worst;

[0016] j. With the goal of prioritizing the selection of operating points near the worst operating point, after selecting operating points from the test operating point set and adding them to the experimental operating point set, return to step e and repeat steps e to h until the prediction accuracy of the turbine blade dynamic stress prediction model on the test set meets the requirements.

[0017] In one or more embodiments, the operating parameters include Mach number, pressure altitude, turbine speed, intake temperature, intake pressure, chamber pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, and fuel flow rate.

[0018] In one or more embodiments, the operating parameters other than the Mach number and the pressure height are obtained through a test run program chart.

[0019] In one or more embodiments, in step c, the set of test operating conditions is obtained by random sampling or stratified random sampling.

[0020] In one or more embodiments, in step d:

[0021] The test condition point located within the boundary of the flight envelope is defined as the test condition point, and the set of test condition points is the set of test condition points;

[0022] Let the ratio of the number of training sets to the number of test sets for the blade dynamic stress prediction model be a:b, where a and b are coprime;

[0023] The initial test condition point is obtained by solving the following formulas (1a) to (1d):

[0024] max[min(|C i -C j |)](1a);

[0025] stC i Cj ∈C exp ∈C line (1b);

[0026] i≠j(1c);

[0027] crad(C exp ) = a + b(1d);

[0028] Among them, C i C j C is the initial test condition point. exp Let C be the set of test operating points. line For the test operating point, C exp This is the set of test operating conditions.

[0029] In one or more embodiments, in step e, the stable operating condition parameters corresponding to each operating point test procedure graph are extracted as input, the maximum vibration stress parameters of the turbine blades under the corresponding parameters are extracted as output, and the test dataset is established based on the input and the output.

[0030] In one or more embodiments, a suitable machine learning model is selected based on the vibration characteristics of the engine blades to train the turbine blade dynamic stress prediction model.

[0031] In one or more embodiments, the turbine blade dynamic stress prediction model is obtained by the following set of formulas (2a) to (2d):

[0032] minf eval (f pre C test (2a);

[0033] stC train ∪C test =C exp (2b);

[0034]

[0035] crad(C train )=m(2d);

[0036] Among them, f eval C is the model evaluation function, where it is assumed that the smaller the evaluation value, the higher the model accuracy. train For the training set of operating points, C test Let m be the set of test operating points, and f be the number of operating points used as the training operating point set. pre This is a model for predicting the dynamic stress of turbine blades.

[0037] In one or more embodiments, the worst-case operating point is obtained by solving for the following formulas (3a) and (3b):

[0038] maxf eval (f pre C worst (3a);

[0039] stC worst ∈C test (3b);

[0040] Among them, f eval C is the model evaluation function. worst For the worst operating condition, f pre This is a model for predicting the dynamic stress of turbine blades.

[0041] In one or more embodiments, step j, with the goal of prioritizing the selection of operating points near the worst operating point, involves selecting operating points from the set of test operating points through the following steps:

[0042] The distance between each operating point in the set of operating points to be tested is calculated using the following formula (4):

[0043] d = min(d tresh ,|C i -C worst |),C i ∈(CC exp (4);

[0044] Where, d tresh This is the distance threshold; when the distance exceeds this value, it is set to d. tresh C is the set of operating conditions to be tested. exp For the test operating condition point set;

[0045] The probability of each working point being selected is calculated using the following formula (5):

[0046]

[0047] After calculating the probability of selecting the operating point, the operating point is selected and added to the test operating point set using the roulette wheel selection method.

[0048] The beneficial effects of this invention are as follows:

[0049] The method for formulating the airworthiness compliance verification test matrix described in one or more of the foregoing embodiments employs a joint machine learning model and an iterative approach using test datasets. The prediction accuracy of the machine learning model on the test set is used as an indicator, and the test dataset is continuously updated until the prediction accuracy meets the indicator. The training set of operating points is then output as the airworthiness compliance verification test matrix for the aero-engine. This ensures that the airworthiness compliance verification test matrix can cover the vibration characteristics of the blades throughout the entire flight envelope. Subsequently, combined with the engine airworthiness compliance verification test, the vibration stress margin of the engine blades within the flight envelope can be effectively evaluated, supporting engine development and airworthiness certification.

[0050] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0051] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0052] Figure 1 A flowchart illustrating some embodiments of the method for developing this airworthiness compliance verification test matrix is ​​shown;

[0053] Figure 2 A schematic diagram of an embodiment of the flight envelope of an aircraft engine is shown;

[0054] Figure 3 A schematic diagram of one embodiment of the test run procedure diagram is shown;

[0055] Figure 4 A schematic diagram showing the distribution of the set of test operating points is provided.

[0056] Figure 5 A schematic diagram showing the distances between operating points is provided.

[0057] Figure 6 This diagram shows the distribution of the test condition point set after the initial test condition points were placed.

[0058] Figure 7A as well as Figure 7B Examples of schematic diagrams showing both the accuracy requirements and those that do not are provided.

[0059] Figure 8This diagram illustrates the process of selecting operating points and adding them to the test operating point set according to an embodiment of this application. Detailed Implementation

[0060] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0061] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0062] Existing test matrix designs rely on experience, and the test data can only illustrate the vibration stress characteristics of the aero-engine at the test point, failing to reflect whether it meets airworthiness requirements across the entire flight envelope. To obtain a test matrix capable of acquiring the vibration characteristics of the aero-engine across its entire flight envelope, according to some embodiments of this application, a method for developing an airworthiness compliance verification test matrix is ​​provided. Figure 1 The diagram illustrates a flowchart of some embodiments of the method for developing this airworthiness compliance verification test matrix, which includes the following steps:

[0063] Step S101: Obtain the flight envelope declared by the engine, for example Figure 2 The diagram shows an embodiment of the flight envelope of an aircraft engine. The aircraft envelope is a closed geometric figure that uses parameters such as flight speed, altitude, overload, and ambient temperature as coordinates to represent the flight range and operating restrictions of an aircraft.

[0064] Step S102: Normalize the parameters according to the range of all operating condition parameters set within the flight envelope.

[0065] In one specific embodiment, the operating parameters will include Mach number, pressure altitude, turbine speed, intake temperature, intake pressure, chamber pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, and fuel flow rate.

[0066] Furthermore, in a specific embodiment, operating parameters other than Mach number and pressure height are obtained through a test procedure graph. Specifically, the variation curves of operating parameters other than Mach number and pressure height can be plotted in the test procedure graph, such as... Figure 3The graph shows the variation of parameters such as turbine speed, intake air temperature, and chamber pressure, from which operating parameters can be obtained.

[0067] Step S103: Obtain the test condition point, such as... Figure 4 As shown, gray dot 1 on the flight envelope diagram of the aero-engine indicates a uniformly set set of test operating condition points C. It can be understood that... Figure 4 The Mach number and pressure height in the data are normalized parameters.

[0068] In a specific embodiment, step S103 involves obtaining the set of test operating points C by random sampling or stratified random sampling.

[0069] Step S104: Obtain the initial test condition point and the set of test condition points, and put the initial test condition point into the set of test condition points.

[0070] In one specific embodiment, in step S104, the boundary of the flight envelope is defined ( Figure 4 The area within the solid line represents test condition point C. line The test operating point C line The test operating point set is constructed, with each operating point containing two parameters: test Mach number and pressure height. Let the ratio of the number of training data points to the number of test data points for the blade dynamic stress prediction model be a:b, where a and b are coprime. Then, the selection of the initial test operating points is transformed into an optimization problem as shown in equations (1a) to (1d):

[0071] max[min(|C i -C j |)](1a);

[0072] stC i C j ∈C exp ∈C line (1b);

[0073] i≠j(1c);

[0074] crad(C exp ) = a + b(1d);

[0075] Among them, C i C j C is the initial test condition point. exp Let be the set of test operating points, where the number of test operating points is a+b. By solving the optimization problem shown in formula (1a), we can obtain the combination of initial test operating points, where the shortest distance between all operating points is greater than that between other combinations.

[0076] Step S105: Based on the test condition point set Cexp Conduct experiments and establish experimental datasets.

[0077] In a specific embodiment, step S105 includes extracting the corresponding stable operating condition parameters from the test procedure graph for each operating point, specifically including: turbine speed, intake air temperature, intake air pressure, chamber pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, fuel flow rate, etc., as well as operating point parameters: Mach number and pressure altitude. These parameters are used as input x, and the maximum vibration stress parameter of the turbine blade under the corresponding parameters is extracted as output y. A test dataset is then established based on the corresponding input x and output y.

[0078] Step S106: Train the turbine blade dynamic stress prediction model.

[0079] In a specific embodiment, step S106 involves selecting a suitable machine learning model based on the engine blade vibration characteristics, such as a generalized linear model, support vector machine, nearest neighbor regression, decision tree, ensemble method, neural network model, etc., to train the turbine blade dynamic stress prediction model f. pre After completing model training, calculate the performance of the prediction model on the test set, such as mean absolute error, root mean square error, root mean square error, normalized root mean square error, and coefficient of determination.

[0080] Step S107: Arrange the test condition point set C according to any scale. exp Extract as a training condition point set C train and the set of test condition points C test And the training condition point set C train The corresponding test dataset of the test procedure diagram is used as the training set, and the set of test condition points C is used as the training set. test The test dataset of the corresponding test procedure diagram is used as the test set.

[0081] In a specific embodiment, step S107 includes selecting m operating points as the training operating point set C. train The dataset is used as the training set, and the remaining n operating conditions are used as the test operating condition set C. test The training set is used as the test set. It's understood that the number of training data points is not *m*, but rather the experimental dataset extracted from the test procedure graphs corresponding to *m* operating points, which is much larger than *m*. Similarly, the number of test data points *n* is not *n*, but rather the experimental dataset extracted from the test procedure graphs corresponding to *n* operating points, which is much larger than *n*. Here, *m* and *n* are any coprime constants.

[0082] In a specific embodiment, the turbine blade dynamic stress prediction model is obtained through the following set of formulas (2a) to (2d):

[0083] minf eval (f pre C test (2a);

[0084] stC train ∪C test =C exp (2b);

[0085]

[0086] crad(C train )=m(2d);

[0087] Among them, f eval Let C be the model evaluation function, where it is assumed that the smaller the evaluation value, the higher the model accuracy. The problem of optimizing the prediction model accuracy is solved according to equation (2a) by continuously adjusting the training working point set C. train The combination of these factors results in the model achieving the highest accuracy when predicting the test set, yielding the turbine blade dynamic stress model f. pre .

[0088] Step S108: Determine the turbine blade dynamic stress prediction model f pre If the prediction accuracy on the test set meets the requirements, the blade dynamic stress prediction model and the training working point set are output; otherwise, proceed to step S109.

[0089] Step S109: Extract the worst-case scenario C from the set of test scenario points, where the model prediction accuracy is the worst. worst ;

[0090] In one specific embodiment, the worst operating condition C worst The following formulas (3a) and (3b) are used to solve for the following:

[0091] maxf eval (f pre C worst (3a);

[0092] stC worst ∈C test (3b);

[0093] Wherein, the model evaluation function f is assumed. eval The smaller the evaluation value, the higher the model accuracy; the worst-case condition C. worst To satisfy the solution of the optimization problem according to formula (3a).

[0094] Step S110: Prioritize selecting the worst operating condition point C. worst Targeting nearby operating points, select operating points from the test operating point set and add them to the experimental operating point set C. expThen, return to step S105 and repeat steps S105 to S108 until the prediction accuracy of the turbine blade dynamic stress prediction model on the test set meets the requirements.

[0095] In one specific embodiment, step S110 prioritizes selecting the worst-case operating condition point C. worst Selecting nearby operating points from the set of test operating points is achieved through the following steps:

[0096] like Figure 5 As shown, black dot 2 represents the worst-case scenario C. worst Gray triangle 3 and square 4 are points in the set of test points C that have not been tested. The distance between each test point i is calculated as follows (4):

[0097] d = min(d tresh ,|C i -C worst |),C i ∈(CC exp (4);

[0098] Where, d tresh This is the distance threshold; when the distance exceeds this value, it is set to d. tresh The probability of each working point being selected is calculated using the following formula (5):

[0099]

[0100] Where, d i The distance C from the operating point is calculated using formula (4). worst The closer the distance, the higher the probability of being selected. When the distance exceeds the threshold d... tresh When the probability of selection is consistent, this ensures that the worst-case operating condition C is selected first. worst Nearby operating points improve the prediction model for turbine blade dynamic stress. pre The accuracy is achieved by uniformly selecting the point C at the worst working condition. worst For points further away, the worst-case scenario C of the model is taken into account. worst and the distance from the worst operating condition point C worst The prediction effect at more distant operating points should be considered to avoid the model failing at the worst operating point C. worst Near-term overfitting. After calculating the probability of selecting the operating point, a+b operating points are selected using a roulette wheel selection method and added to the test operating point set C. exp Using the aforementioned method, at the test condition point set C expDuring the addition of operating points, the reciprocal of the distance between operating points is used as a probability metric for selecting an operating point, and a distance threshold is set. This approach balances the advantages of prioritizing the selection of new operating points near the operating point with the worst prediction accuracy, and selecting points farther from the worst operating point with equal probability. Combined with the roulette wheel selection method, model overfitting is avoided. By gradually adding operating points based on the initial set, the prediction accuracy of blade dynamic stress is improved. While meeting accuracy requirements, a reasonable airworthiness compliance verification matrix is ​​formulated to minimize the number of required operating points, avoiding redundancy in airworthiness compliance verification test operating points, saving test costs, and reducing the test cycle. Roulette wheel selection: This is a proportional selection operation. The probability distribution is viewed as a roulette wheel, and the size of each slice is proportional to the individual's selection probability. The selection process is like spinning the roulette wheel; when it stops spinning, the individual corresponding to the top slice is selected. Here, a and b in a+b are any coprime constants.

[0101] In some specific embodiments, in step S110, specifically, the set of test operating points C... exp Experiments were conducted on operating points that had not yet been tested, and the results were extracted to supplement the experimental dataset. Then, a training set C of operating points was randomly selected according to an m:n ratio. train and the set of test condition points C test Training a turbine blade dynamic stress prediction model f pre By adjusting the training condition point set C train The combination yields the turbine blade dynamic stress prediction model f with the best prediction accuracy on the test set. pre To determine if the prediction accuracy meets the requirements, if not, select a+b more operating points to expand the test dataset using the method described above, until the turbine blade dynamic stress prediction model f is obtained. pre To meet accuracy requirements, output the turbine blade dynamic stress prediction model f. pre And the training set operating conditions, where the training set operating condition points are the test matrix, can be used for airworthiness compliance verification tests of aero-engines. Here, a and b in a+b are arbitrarily coprime constants, and m and n in m:n are arbitrarily coprime constants.

[0102] The method for formulating the airworthiness compliance verification test matrix described in one or more of the foregoing embodiments employs a joint machine learning model and an iterative approach using test datasets. The prediction accuracy of the machine learning model on the test set is used as an indicator, and the test dataset is continuously updated until the prediction accuracy meets the indicator. The training set of operating points is then output as the airworthiness compliance verification test matrix for the aero-engine. This ensures that the airworthiness compliance verification test matrix can cover the vibration characteristics of the blades throughout the entire flight envelope. Subsequently, combined with the engine airworthiness compliance verification test, the vibration stress margin of the engine blades within the flight envelope can be effectively evaluated, supporting engine development and airworthiness certification.

[0103] The present invention is further illustrated by the following specific embodiment:

[0104] First, input the flight envelope declared by the engine. Then, normalize the operating parameters according to the range of all operating parameters set within the flight envelope. The operating parameters will include Mach number, pressure altitude, turbine speed, inlet temperature, inlet pressure, cabin pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, fuel flow rate, etc.

[0105] Subsequently, a set of test condition points C was uniformly set on the flight envelope diagram of the aero-engine.

[0106] Subsequently, the number of training sets and test sets for the turbine blade dynamic stress prediction model were set at a ratio of 3:1. By solving the optimization problem shown in formula (1a), four initial test conditions were obtained and placed into the test condition set C. exp In the middle, as attached Figure 6 The black dot 5 is shown in the image.

[0107] Subsequently, the set of test operating points C was extracted. exp Tests were conducted at operating points not previously tested. For each operating point, the corresponding stable turbine speed, intake air temperature, intake air pressure, chamber pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, and fuel flow rate were obtained from the test procedure diagrams. Additionally, the operating point parameters, Mach number and pressure altitude, were also acquired. These parameters were used as x, and the maximum turbine blade vibration stress parameter under the corresponding parameters was extracted as y, thus establishing the test dataset.

[0108] Subsequently, the test condition point set C was selected at a ratio of 3:1. exp Extract the training condition point set C from . train and the set of test condition points C test The training condition point set C train The corresponding test dataset of the test procedure diagram is used as the training set, and the set of test condition points C is used as the training set. testThe test dataset of the corresponding test procedure diagram is used as the test set, and the blade dynamic stress prediction model f is trained based on the decision tree model. pre And calculate the model's accuracy on the test set. This is achieved by continuously adjusting the training condition point set C. train Solving the optimization problem shown in equation (2a), we obtain the blade dynamic stress prediction model f with the best accuracy on the test set. pre .

[0109] Subsequently, the dynamic stress prediction model f of the blade was determined. pre In this embodiment, the prediction accuracy on the test set is considered to meet the requirement if it meets 0.3 times the dispersion band. For example, when obtaining... Figure 7A If the result shown does not meet the requirements, proceed to the next step. When you obtain, for example... Figure 7B If the result meets the requirements, then execute the output step.

[0110] Subsequently, by solving the optimization problem shown in equation (3a), the set of test condition points C is obtained. test The operating condition C with the worst prediction accuracy in the model worst .

[0111] The set C of test operating points that were not selected is calculated according to formula (5). exp The probability of selecting the working condition points is calculated, and four working condition points are selected and added to the test working condition point set C using a roulette wheel selection method. exp As attached Figure 8 The black outline of the gray dot 6, then return to the experimental steps.

[0112] Once the prediction accuracy of the blade dynamic stress prediction model on the test set meets the requirements (in this embodiment, the predicted dynamic stresses are all within 0.3 times the dispersion band), the blade dynamic stress prediction model f is output. pre The set of training test points C corresponding to the model train This set of operating points is the airworthiness compliance test matrix generated by this invention.

[0113] In the airworthiness compliance verification test of aero-engines, the test is carried out directly based on the test matrix obtained from the above steps, and the test data is extracted to train the blade dynamic stress prediction model f under the same form. pre Based on the model's prediction of the maximum dynamic stress point of the turbine blades within the operating parameter range, it is determined whether the engine meets airworthiness requirements.

[0114] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any way. This application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for developing an airworthiness compliance verification test matrix, characterized in that, Includes the following steps: a. Obtain the flight envelope declared by the engine; b. Normalize the parameters according to the range set for all operating condition parameters within the flight envelope; c. Obtain the test condition point; d. Obtain the initial test condition point and the set of test condition points, and put the initial test condition point into the set of test condition points; e. Conduct tests based on the set of test operating points and establish a test dataset; f. Train the turbine blade dynamic stress prediction model; g. Extract the test condition point set into a training condition point set and a test condition point set according to any ratio, and use the test dataset of the test program graph corresponding to the training condition point set as the training set, and use the test dataset of the test program graph corresponding to the test condition point set as the test set. h. Determine whether the prediction accuracy of the turbine blade dynamic stress prediction model on the test set meets the requirements. If it does, output the blade dynamic stress prediction model and the training operating point set. If it does not meet the requirements, proceed to step i. i. Extract the worst-case scenario point from the set of test scenario points where the model prediction accuracy is the worst; j. With the goal of prioritizing the selection of operating points near the worst operating point, after selecting operating points from the test operating point set and adding them to the experimental operating point set, return to step e and repeat steps e to h until the prediction accuracy of the turbine blade dynamic stress prediction model on the test set meets the requirements.

2. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, The operating parameters include Mach number, pressure altitude, turbine speed, intake temperature, intake pressure, chamber pressure, ambient pressure, compressor adjustable stator blade angle, transition bleed air valve angle, and fuel flow rate.

3. The method for developing the airworthiness compliance verification test matrix as described in claim 2, characterized in that, The operating parameters, excluding the Mach number and pressure height, are obtained through the test run program diagram.

4. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, In step c, the set of test operating points is obtained by random sampling or stratified random sampling.

5. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, In step d: The test condition point located within the boundary of the flight envelope is defined as the test condition point, and the set of test condition points is the set of test condition points; Let the ratio of the number of training sets to the number of test sets for the blade dynamic stress prediction model be a:b, where a and b are coprime; The initial test condition point is obtained by solving the following formulas (1a) to (1d): max[min(|C i -C j |)](1a); s.t.C i ,C j ∈C exp ∈C line (1b); i≠j(1c); crad(C exp )=a+b(1d); Among them, C i C j C is the initial test condition point. exp Let C be the set of test operating points. line For the test operating point, C exp This is the set of test operating conditions.

6. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, In step e, the stable operating condition parameters corresponding to each operating point test program graph are extracted as input, and the maximum vibration stress parameters of the turbine blades under the corresponding parameters are extracted as output. The test dataset is established based on the input and the output.

7. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, The turbine blade dynamic stress prediction model is trained by selecting an appropriate machine learning model based on the vibration characteristics of the engine blades.

8. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, The prediction model for dynamic stress in turbine blades is obtained through the following set of formulas (2a) to (2d): minf eval (f pre ,C test )(2a); s.t.C train ∪C test =C exp (2b); crad(C train )=m(2d); Among them, f eval C is the model evaluation function, where it is assumed that the smaller the evaluation value, the higher the model accuracy. train For the training set of operating points, C test Let m be the set of test operating points, and f be the number of operating points used as the training operating point set. pre This is a model for predicting the dynamic stress of turbine blades.

9. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, The worst operating condition point is obtained by solving the following formulas (3a) and (3b): maxf eval (f pre C worst (3a) s.t.C worst ∈C test (3b); Among them, f eval C is the model evaluation function. worst For the worst operating condition, f pre This is a model for predicting the dynamic stress of turbine blades.

10. The method for developing the airworthiness compliance verification test matrix as described in claim 1, characterized in that, In step j, with the goal of prioritizing the selection of operating points near the worst operating point, the selection of operating points from the test operating point set is achieved through the following steps: The distance between each operating point in the set of operating points to be tested is calculated using the following formula (4): d=min(d tresh ,|C i -C worst |),C i ∈(C-C exp ) (4); Where, d tresh This is the distance threshold; when the distance exceeds this value, it is set to d. tresh C is the set of operating conditions to be tested. exp For the test operating condition point set; The probability of each working point being selected is calculated using the following formula (5): After calculating the probability of selecting the operating point, the operating point is selected and added to the test operating point set using the roulette wheel selection method.

Images