Aero-engine comprehensive performance evaluation method based on test run data and network toughness

Abstract

The invention discloses an aero-engine comprehensive performance evaluation method based on test run data and network toughness, and the method comprises the steps: obtaining the test run data for a plurality of aero-engines of the same model, which are qualified in test run under preset test run parameters, and carrying out the monitoring index screening; constructing a performance network model and a performance dynamics model corresponding to the aero-engine; utilizing a dynamic dimension reduction method to construct comprehensive performance indexes of the aero-engine; determining each kinetic parameter of each aero-engine based on the test run data so as to construct a qualified value range of each kinetic parameter; based on the qualified value range of each kinetic parameter, determining a qualified interval of the comprehensive performance of the aero-engine by using the comprehensive performance index of the aero-engine; and for the aero-engine to be subjected to comprehensive performance evaluation, acquiring test run data of the aero-engine, calculating a corresponding comprehensive performance index of the aero-engine, and judging whether the comprehensive performance evaluation of the aero-engine is qualified or not according to the qualified interval.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft engine performance evaluation, and in particular to an aircraft engine comprehensive performance evaluation method based on test data and network resilience. Background Art

[0002] Aircraft engines are highly complex and sophisticated thermomechanical systems with high technology concentration and content, complex system sets, high added value, and strong driving force. They are the top priority of the national industrial development strategy. Aircraft engine components are complicated, operating conditions are diverse, and the working environment is complex. Their comprehensive performance evaluation faces huge challenges. Sudden failures will greatly affect the performance of aircraft engines, causing huge economic losses, and will seriously threaten the safety of personnel lives, and even directly endanger national security. Accurate performance evaluation can reflect the performance status and state change rules of the equipment, minimize downtime caused by performance degradation, facilitate active maintenance, and prevent huge losses caused by accidents. Therefore, carrying out comprehensive performance evaluation of aircraft engines is the core basic work of aircraft engine health management, and is an important guarantee for the safe and efficient operation of aircraft engines. It can further improve the intelligent operation level of aircraft engines, highlight core competitiveness, and optimize resource allocation.

[0003] In actual engineering applications, the performance evaluation of aircraft engines has been widely carried out. Airlines usually rely on a single parameter, such as exhaust gas temperature (EGT) or EGT margin to evaluate the performance of aircraft engines; aircraft engine manufacturers have a complete set of quality inspection and test procedures to evaluate the condition of aircraft engines, focusing on the performance of aircraft engines under several typical conditions, by setting different thrusts, high-pressure compressor inlet guide vane angles, and high-pressure compressor speeds, simulating several typical conditions, focusing on speed, thrust, temperature and other indicators and their interrelationships, to verify whether they meet the requirements.

[0004] However, both of the above methods have obvious defects: the single parameter method used by airlines is easy to operate and can quickly provide a certain basis for maintenance decisions, but the aircraft engine is a typical complex system with many parameters affecting its performance. Relying on a single parameter alone cannot fully reflect its actual performance; the quality inspection and test procedures used by aircraft engine manufacturers can help manufacturers conduct basic inspections of aircraft engine performance, but considering certain key monitoring indicators and their interrelationships in isolation cannot represent its comprehensive performance. In addition, in actual use, the working state of aircraft engines is constantly changing, and it is incomplete to evaluate the performance under only a few typical states. Therefore, in view of the above defects, it is necessary to study how to improve the comprehensive performance evaluation capabilities of aircraft engines in actual engineering.

[0005] Existing technologies for aircraft engine performance evaluation can be roughly divided into two categories: first, traditional system comprehensive evaluation technology and its improved technology; second, artificial intelligence-based technology.

[0006] For the first type of solution, for example, the patent application with publication number CN118535869A: An aircraft engine evaluation method, device and system, guided by the "decomposition-integration" concept, aims to conduct comprehensive performance evaluation of aircraft engines, a typical complex system, in a big data environment. However, the shortcomings of this type of solution are that the independence of indicators will affect the accuracy of weight calculation, and the evaluation results are mainly obtained by linear operation of indicators, which cannot reflect the qualitative changes and emergent laws of the system caused by the nonlinear relationship and fusion of indicators, making it difficult to accurately evaluate the state of complex systems, and may even make an incorrect estimate of the system state.

[0007] For the second type of solution, for example, the patent application with publication number CN114997051A: a method for aircraft engine life prediction and health assessment based on transfer learning, which uses a neural network model to predict the life and health assessment of aircraft engines; however, this type of artificial intelligence-based technology has a black box characteristic, which means that it is impossible to clearly understand the result generation mechanism of the model, and it is difficult to scientifically explain the mechanism of aircraft engine state changes and their internal relationships. Summary of the invention

[0008] The purpose of the present invention is to provide an aero-engine comprehensive performance evaluation method based on test data and network resilience, so as to overcome the problems existing in the existing aero-engine performance evaluation methods.

[0009] In order to achieve the above tasks, the present invention adopts the following technical solutions:

[0010] A comprehensive performance evaluation method for aircraft engines based on test data and network resilience, including:

[0011] Step 1, for multiple aircraft engines of the same model that have passed the test under preset test parameters, obtain test data and perform monitoring index screening;

[0012] Step 2: for each selected monitoring indicator of an aircraft engine, a performance network model of the corresponding aircraft engine is constructed based on complex network theory and the relationship between the monitoring indicators;

[0013] Step 3, introducing the preset test parameters into the gene regulation dynamics and performing fitting, so as to construct a performance dynamics model of each aircraft engine;

[0014] Step 4, for the performance network model of the aircraft engine, a dynamic dimension reduction method is used to construct a comprehensive performance index of the aircraft engine;

[0015] Step 5, based on the test data, using the performance dynamics model and the performance network model of the aircraft engine to determine the dynamic parameters of each aircraft engine, so as to construct the qualified value range of each dynamic parameter;

[0016] Step 6, based on the qualified value ranges of the dynamic parameters, using the comprehensive performance index of the aircraft engine to determine the qualified interval of the comprehensive performance of the aircraft engine;

[0017] Step 7, for the aircraft engine to be subjected to comprehensive performance evaluation, a test run is carried out under preset parameters to obtain the test run data of the aircraft engine, and after screening the monitoring indicators in the test run data, a performance network model and a performance dynamics model are constructed, the dynamic parameters are determined and the corresponding comprehensive performance indicators of the aircraft engine are calculated; based on the relationship between the indicators and the qualified range of the comprehensive performance of the aircraft engine, it is determined whether the comprehensive performance evaluation of the aircraft engine is qualified.

[0018] Furthermore, the monitoring indicator screening refers to taking the difference of all monitoring indicators and calculating the corresponding variance, and eliminating the monitoring indicators with zero variance.

[0019] Furthermore, the performance network model of the corresponding aircraft engine is constructed based on the complex network theory and the relationship between the monitoring indicators, including:

[0020] The monitoring indicators in the aircraft engine test data are abstracted into nodes, and the relationships between the monitoring indicators are abstracted into edges between nodes, so as to construct a performance network model for each aircraft engine; the relationships between the monitoring indicators are represented by correlation coefficients, which are expressed as follows:

[0021] ;

[0022] In the above formula, It represents the i-th and j-th monitoring indicators among the N monitoring indicators selected for aircraft engines; There are two monitoring indicators The difference and The corresponding covariance is and are the differences of the monitoring indicators. and The corresponding standard deviation;

[0023] The weighted adjacency matrix is ​​used to construct the edges between nodes, which is expressed as follows:

[0024] ;

[0025] in, is the weighted adjacency matrix The element in row i and column j in Construct a threshold for the network model, ;when When and There is an edge between them, otherwise there is no edge; when two monitoring indicators and Correlation coefficient When it is positive, it is positively correlated, recorded as ; Negative value indicates negative correlation, recorded as .

[0026] Furthermore, the preset test parameters are introduced into the gene regulation dynamics and fitted, so as to construct the performance dynamics model of each aircraft engine, which is specifically expressed as:

[0027] ;

[0028] In the formula , represents the i-th and j-th monitoring indicators in the aircraft engine performance network model, For monitoring indicators The proportion of the normal range of the preset indicators corresponding to the regression, is the number of isolated nodes without edges in the aircraft engine performance network model; , Respectively represent monitoring indicators The excitation intensity and suppression intensity under the thrust and IGV angle of the preset test parameters, where , They are , The positive part of , They are , The negative part of and represents the overall excitation and inhibition terms, and .

[0029] Furthermore, the performance network model for the aircraft engine uses a dynamic dimension reduction method to construct an aircraft engine comprehensive performance index, including:

[0030] According to the mean field theory, the dynamic equations of all nodes in the performance dynamics model of the aircraft engine are coupled into one-dimensional equations and the expectation is calculated to obtain the average state of the complex system of the aircraft engine, and further derive the relationship between the average behavior and structure of the system:

[0031] ;

[0032] in, , , and They are , , and average value; is the average weighted degree of the performance network model, through the weighted adjacency matrix Calculated; is the proportion of negatively correlated edges between monitoring indicators in the performance network model;

[0033] The above formula (5) is changed into a matrix multiplication form, which is expressed as follows:

[0034] ;

[0035] use and Representing the above two matrices, let:

[0036] , ;

[0037] Based on the Frobenius norm, use To measure coupling performance, use The inverse of is used to measure the activity performance, and the comprehensive performance index of the aircraft engine is as follows:

[0038] .

[0039] Furthermore, based on the test data, the performance dynamics model and the performance network model of the aircraft engine are used to determine the dynamic parameters of each aircraft engine, thereby constructing the qualified value range of each dynamic parameter, including:

[0040] Through the performance dynamics model and performance network model, the dynamic parameters of each aircraft engine are obtained. , , , The minimum and maximum values ​​of the dynamic parameters of all aircraft engines constitute the qualified value range of the corresponding dynamic parameters.

[0041] Furthermore, based on the qualified value ranges of the dynamic parameters, the qualified interval of the comprehensive performance of the aircraft engine is determined by using the comprehensive performance index of the aircraft engine, including:

[0042] By taking values ​​within the qualified range of values ​​of each dynamic parameter and substituting them into the comprehensive performance index of the aero-engine, a series of coupling performance and activity performance are obtained; by constructing a two-dimensional coordinate diagram of the coupling performance and activity performance, the area formed by the corresponding coordinate points of all coupling performance and activity performance in the coordinate diagram is the qualified interval of the comprehensive performance of the aero-engine.

[0043] An aircraft engine comprehensive performance evaluation device comprises a processor, a memory and a computer program stored in the memory; when the processor is executed by a computer, the aircraft engine comprehensive performance evaluation method based on test data and network resilience is implemented.

[0044] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the method for evaluating the comprehensive performance of an aircraft engine based on test data and network resilience is implemented.

[0045] Compared with the prior art, the present invention has the following technical features:

[0046] 1. The present invention is based on test data and complex networks and can objectively and globally evaluate highly integrated, internally coupled and nonlinear aircraft engine systems based on real aircraft engine test data.

[0047] 2. The aviation engine performance network model and aviation engine performance dynamics model constructed by the present invention can accurately describe the state of the system at each stage and explain the relationship between the monitoring indicators and the system state change pattern, and are interpretable.

[0048] 3. The comprehensive performance evaluation index of the aircraft engine proposed in the present invention can dynamically characterize the status of the aircraft engine at each stage, distinguish between qualified and faulty aircraft engines, and provide guidance for production. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the process of the present invention;

[0050] Figure 2 Schematic diagram of an aircraft engine performance network model in an embodiment of the present invention;

[0051] Figure 3 is a fitting error diagram of the aero-engine performance dynamics model in an embodiment of the present invention;

[0052] Figure 4 is a dynamic evaluation result of the comprehensive performance of the aircraft engine in an embodiment of the present invention;

[0053] Figure 5 It is the qualified range of the comprehensive performance of the aircraft engine obtained in the embodiment of the present invention. DETAILED DESCRIPTION

[0054] Aiming at the test data of aircraft engines, the present invention combines the advantages of complex networks and network resilience theory, and proposes an aircraft engine comprehensive performance evaluation method based on test data and complex network resilience, which can accurately and dynamically characterize the comprehensive performance of aircraft engines, obtain the comprehensive performance range corresponding to qualified aircraft engines, and use this as the judgment standard for aircraft engine fault diagnosis.

[0055] See attached Figure 1 The present invention provides an aero-engine comprehensive performance evaluation method based on test data and network resilience, comprising the following steps:

[0056] Step 1: for multiple aircraft engines of the same model that have passed the test under preset test parameters, obtain test data and screen monitoring indicators; wherein the preset test parameters include thrust and IGV angle.

[0057] The test data of aircraft engines is composed of multiple monitoring indicators, such as low-pressure rotor speed, lubricating oil temperature, etc. This scheme obtains the corresponding test data for the same type of aircraft engines that have passed the test, takes the difference of all monitoring indicators and calculates the corresponding variance, and eliminates the monitoring indicators with zero variance, thereby completing the screening of monitoring indicators. It should be noted that the test data of aircraft engines that have passed the test are collected under the same preset thrust and IGV angle.

[0058] Step 2: For each aircraft engine's selected monitoring indicators, a performance network model of the corresponding aircraft engine is constructed based on complex network theory and the relationship between the monitoring indicators.

[0059] Based on complex network theory, the monitoring indicators in the aircraft engine test data are abstracted into nodes, and the relationships between the monitoring indicators are abstracted into edges between nodes, so as to construct a performance network model for each aircraft engine.

[0060] The correlation coefficients between the monitoring indicators are used to quantify the strength of their correlations, which are expressed as follows:

[0061] (1);

[0062] in, It represents the i-th and j-th monitoring indicators among the N monitoring indicators selected for aircraft engines; There are two monitoring indicators The difference and The corresponding covariance is and are the differences of the monitoring indicators. and The corresponding standard deviation.

[0063] The weighted adjacency matrix is ​​used to construct the edges between nodes, which is expressed as follows:

[0064] (2);

[0065] in, is the weighted adjacency matrix The element in row i and column j in Construct a threshold for the network model, ;when When and There is an edge between them, otherwise there is no edge; when two monitoring indicators and Correlation coefficient When it is positive, it is positively correlated, recorded as ; Negative value indicates negative correlation, recorded as ; Based on this, the aircraft engine performance network model is established, see Figure 2 The example given.

[0066] Step 3: introducing the preset test parameters into the gene regulation dynamics and fitting them, so as to construct a performance dynamics model of each aircraft engine.

[0067] Since the positive and negative correlations between monitoring indicators are common in the constructed network model, differential equations are used to simulate the dynamic changes between monitoring indicators; the Michaelis-Menten equation is used to represent gene regulatory interactions, and the original equation form of the gene regulation dynamics is:

[0068] (3);

[0069] Where t is the time parameter, is the Hill coefficient, which is 1 in this scheme; is the protein inactivation rate, and Represent the inter-gene excitation strength and inhibition strength respectively, and The incentive matrix and the inhibition matrix The element in the i-th row and j-th column in .

[0070] Based on gene regulation dynamics, the test parameters that determine the state of the aircraft engine during the test run of the aircraft engine are introduced to construct the performance dynamics model of the aircraft engine as follows:

[0071] (4);

[0072] In the formula , represents the i-th and j-th monitoring indicators (nodes) in the aircraft engine performance network model, For monitoring indicators The proportion of the normal range of the preset indicators corresponding to the regression, is the number of isolated nodes without edges in the aircraft engine performance network model; , Respectively represent monitoring indicators The excitation intensity and suppression intensity under the thrust and IGV angle of the preset test parameters, where , They are , The positive part of , They are , The negative part in ; to simplify the representation, use and represents the overall excitation and inhibition terms, and .

[0073] Among them, here, , The monitoring indicators can be calculated The correlation coefficient with the thrust time series and IGV angle time series is obtained, and then , , and The value of The stable state can be solved by equation (4): get.

[0074] Step 4: Based on the performance network model of the aircraft engine, the dynamic dimension reduction method is used to construct the comprehensive performance index of the aircraft engine.

[0075] According to the mean field theory, the dynamic equations of all nodes in equation (4) are coupled into one-dimensional equations and the expectation is calculated to obtain the average state of the complex system of the aircraft engine, and further derive the relationship between the average behavior and structure of the system:

[0076] (5);

[0077] in, , , and They are , , and average value; is the average weighted degree of the performance network model, through the weighted adjacency matrix Calculated; It is the proportion of negatively correlated edges between monitoring indicators in the performance network model.

[0078] Convert equation (5) into matrix multiplication form, as follows:

[0079] (6);

[0080] Here we use and Represent the above two matrices, that is, let:

[0081] , (7);

[0082] Based on the Frobenius norm, use To measure coupling performance, use The inverse of is used to measure the activity performance, and the comprehensive performance index of the aircraft engine is as follows:

[0083] (8).

[0084] Step 5: Based on the test data, the performance dynamics model and the performance network model of the aircraft engine are used to determine the dynamic parameters of each aircraft engine, so as to construct the qualified value range of each dynamic parameter.

[0085] Through the performance dynamics model in step 3 and the performance network model in step 2, the dynamic parameters of each aircraft engine can be obtained. , , , The minimum and maximum values ​​of the dynamic parameters of all aircraft engines constitute the qualified value range of the corresponding dynamic parameters.

[0086] For example, among all the dynamic parameters of aircraft engines The minimum value of , the maximum value is , then the kinetic parameters The qualified value range of .

[0087] Step 6: Based on the qualified value ranges of the dynamic parameters, the qualified interval of the comprehensive performance of the aircraft engine is determined using the comprehensive performance index of the aircraft engine.

[0088] By taking values ​​from the qualified range of values ​​of each dynamic parameter and substituting them into the comprehensive performance index of the aircraft engine, a series of coupling performance and activity performance can be obtained; by constructing a two-dimensional coordinate graph of the coupling performance and activity performance, the area formed by the corresponding coordinate points of all coupling performance and activity performance in the coordinate graph is the qualified range of the comprehensive performance of the aircraft engine.

[0089] Step 7, for the aircraft engine to be subjected to comprehensive performance evaluation, a test run is carried out under preset parameters to obtain the test run data of the aircraft engine, and after screening the monitoring indicators in the test run data, a performance network model and a performance dynamics model are constructed, the dynamic parameters are determined and the corresponding comprehensive performance indicators of the aircraft engine are calculated; based on the relationship between the indicators and the qualified range of the comprehensive performance of the aircraft engine, it is determined whether the comprehensive performance evaluation of the aircraft engine is qualified.

[0090] That is, when the coupling performance and activity performance of the aircraft engine to be subjected to comprehensive performance evaluation obtained by calculation fall within the qualified range of the comprehensive performance of the aircraft engine described in step 6 at the corresponding coordinate points in the coordinate diagram, the comprehensive performance evaluation result of the aircraft engine is considered qualified, otherwise it is unqualified.

[0091] One embodiment of the present invention takes a certain aircraft engine manufacturer as the research object, and collects the test data of 17 aircraft engines of the same model under preset test parameters, including 13 aircraft engines that passed the test and 4 aircraft engines that failed the test. The aircraft engine test data records dozens of monitoring indicators such as thrust, air pressure and temperature, fuel pressure and temperature, air flow, fuel flow, engine speed, control valve opening, etc. during the aircraft engine test. The data is high-dimensional, heterogeneous, multivariate and multimodal, and the correlation between monitoring indicators is strong.

[0092] By screening the monitoring indicators of the test data of 17 aircraft engines that have passed the test, a total of 28 monitoring indicators are extracted for each aircraft engine, as follows:

[0093] Low-pressure rotor speed (N1), high-pressure rotor speed (N2), turbine after-gas temperature (N3), lubricating oil temperature (N4), main oil circuit pressure (N5), auxiliary oil circuit pressure (N6), high-pressure compressor and low-pressure compressor outlet air pressure ratio (N7), cam box position (N8), fuel low flow sensor (N9), intake casing vibration value (N10), fuel nozzle vibration value (N11), high-pressure turbine bearing vibration value (N12), low-pressure compressor outlet pressure (N13), high-pressure compressor outlet pressure (N14), low-pressure compressor outlet temperature (N15), high-pressure compressor outlet left temperature (N16) , high-pressure compressor outlet right side temperature (N17), test chamber front static pressure (N18), high-pressure compressor and low-pressure turbine outlet air pressure ratio (N19), fuel control voltmeter (N20), catalytic igniter fuel supply pressure (N21), low-pressure fuel pump after-pump pressure (N22), high-pressure fuel pump after-pump pressure (N23), low-pressure compressor outlet air pressure entering the pressure ratio regulator (N24), high-pressure compressor outlet air pressure entering the pressure ratio regulator (N25), low-pressure turbine outlet air pressure entering the pressure ratio regulator (N26), exhaust mixer static pressure (N27), low-pressure fuel pump pump front pressure (N28).

[0094] In this embodiment, the performance network model constructed by the monitoring indicators selected for one of the aircraft engines that passed the test run is as follows: Figure 2 shown.

[0095] In order to evaluate the performance dynamics model constructed by the present invention, this embodiment uses the relationship between the monitoring indicators to obtain the given time t. and And the weighted adjacency matrix , the monitoring index can be calculated by formula (4) The value of is taken as the simulation value, and the error between the simulation value and the true value of the monitoring indicator is calculated, such as Figure 3 As shown; verification shows that there is little difference between the actual value of each monitoring indicator and the simulated value calculated by the performance dynamics model. The mean relative error (MRE) provided in the upper right corner is less than 1%, indicating that the dynamics model constructed with the actual monitoring indicator data can accurately characterize the status of each monitoring indicator and the overall state of the aircraft engine, and the performance dynamics model is very suitable for the analysis of the aircraft engine test process.

[0096] The kinetic parameters in this example , , , The qualified value ranges are: [3.6818, 11.676], [1.2093, 1.8718], [0, 0.15424], [0, 0.3788].

[0097] like Figure 4 As shown, the solution results of the comprehensive performance indicators of the aircraft engines corresponding to the dynamic parameters of the aircraft engines that passed and failed the test in this embodiment are displayed.

[0098] like Figure 5 As shown, Figure 5 The middle green area is the qualified range of the comprehensive performance of the aircraft engine. It can be seen from the figure that the coordinate points corresponding to the comprehensive performance indicators of the 13 aircraft engines that passed the test in this embodiment all fall within the qualified range, while the coordinate points corresponding to the 4 aircraft engines that failed the test are all outside the qualified range, indicating the effectiveness of the evaluation method of the present invention.

[0099] The embodiments show that the aircraft engine performance network model and the aircraft engine performance dynamics model constructed by the present invention have a mean relative error (MRE) of less than 1%, which means that the method of the present invention can accurately describe the state of the system at each stage, and the form of differential equations can explain the relationship between the monitoring indicators and the system state change pattern, which is interpretable; the aircraft engine comprehensive performance evaluation index proposed by the present invention can dynamically characterize the state of the aircraft engine at each stage, distinguish between qualified and faulty aircraft engines, and provide guidance for production, which has obvious advantages.

[0100] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for comprehensive performance evaluation of aircraft engines based on test data and network resilience, characterized in that: include: Step 1, for multiple aircraft engines of the same model that have passed the test under preset test parameters, obtain test data and perform monitoring index screening; Step 2: for each selected monitoring indicator of an aircraft engine, a performance network model of the corresponding aircraft engine is constructed based on complex network theory and the relationship between the monitoring indicators; Step 3, introducing the preset test parameters into the gene regulation dynamics and performing fitting, so as to construct a performance dynamics model of each aircraft engine; Step 4, for the performance network model of the aircraft engine, a dynamic dimension reduction method is used to construct a comprehensive performance index of the aircraft engine; Step 5, based on the test data, using the performance dynamics model and the performance network model of the aircraft engine to determine the dynamic parameters of each aircraft engine, so as to construct the qualified value range of each dynamic parameter; Step 6, based on the qualified value ranges of the dynamic parameters, using the comprehensive performance index of the aircraft engine to determine the qualified interval of the comprehensive performance of the aircraft engine; Step 7, for the aircraft engine to be subjected to comprehensive performance evaluation, a test run is carried out under preset parameters to obtain the test run data of the aircraft engine, and after screening the monitoring indicators in the test run data, a performance network model and a performance dynamics model are constructed, the dynamic parameters are determined and the corresponding comprehensive performance indicators of the aircraft engine are calculated; based on the relationship between the indicators and the qualified range of the comprehensive performance of the aircraft engine, it is determined whether the comprehensive performance evaluation of the aircraft engine is qualified.

2. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 1, characterized in that: The monitoring indicator screening refers to taking the difference of all monitoring indicators and calculating the corresponding variance, and eliminating the monitoring indicators with zero variance.

3. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 1, characterized in that: The performance network model of the corresponding aircraft engine is constructed based on complex network theory and the relationship between monitoring indicators, including: The monitoring indicators in the aircraft engine test data are abstracted into nodes, and the relationships between the monitoring indicators are abstracted into edges between nodes, so as to construct a performance network model for each aircraft engine; the relationships between the monitoring indicators are represented by correlation coefficients, which are expressed as follows: ； In the above formula, It represents the i-th and j-th monitoring indicators among the N monitoring indicators selected for aircraft engines; There are two monitoring indicators The difference and The corresponding covariance is and are the differences of the monitoring indicators. and The corresponding standard deviation; The weighted adjacency matrix is ​​used to construct the edges between nodes, which is expressed as follows: ； in, is the weighted adjacency matrix The element in row i and column j in Construct a threshold for the network model, ;when When and There is an edge between them, otherwise there is no edge; when two monitoring indicators and Correlation coefficient When it is positive, it is positively correlated, recorded as ; Negative value indicates negative correlation, recorded as .

4. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 2, characterized in that: The preset test parameters are introduced into the gene regulation dynamics and fitted to construct the performance dynamics model of each aircraft engine, which is specifically expressed as: ； In the formula , represents the i-th and j-th monitoring indicators in the aircraft engine performance network model, For monitoring indicators The proportion of the normal range of the preset indicators corresponding to the regression, is the number of isolated nodes without edges in the aircraft engine performance network model; , Respectively represent monitoring indicators The excitation intensity and suppression intensity under the thrust and IGV angle of the preset test parameters, where , They are , The positive part of , They are , The negative part of and represents the overall excitation and inhibition terms, and .

5. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 4 is characterized in that: The performance network model for the aircraft engine uses a dynamic dimension reduction method to construct the comprehensive performance index of the aircraft engine, including: According to the mean field theory, the dynamic equations of all nodes in the performance dynamics model of the aircraft engine are coupled into one-dimensional equations and the expectation is calculated to obtain the average state of the complex system of the aircraft engine, and further derive the relationship between the average behavior and structure of the system: ； in, , , and They are , , and average value; is the average weighted degree of the performance network model, through the weighted adjacency matrix Calculated; is the proportion of negatively correlated edges between monitoring indicators in the performance network model; The above formula (5) is changed into a matrix multiplication form, which is expressed as follows: ； use and Representing the above two matrices, let: ， ； Based on the Frobenius norm, use To measure coupling performance, use The inverse of is used to measure the activity performance, and the comprehensive performance index of the aircraft engine is as follows: 。 6. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 5, characterized in that: Based on the test data, the performance dynamics model and the performance network model of the aircraft engine are used to determine the dynamic parameters of each aircraft engine, thereby constructing the qualified value range of each dynamic parameter, including: Through the performance dynamics model and performance network model, the dynamic parameters of each aircraft engine are obtained. , , , The minimum and maximum values ​​of the dynamic parameters of all aircraft engines constitute the qualified value range of the corresponding dynamic parameters.

7. The method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to claim 1, characterized in that: Based on the qualified value ranges of the dynamic parameters, the qualified interval of the comprehensive performance of the aircraft engine is determined by using the comprehensive performance index of the aircraft engine, including: By taking values ​​within the qualified range of values ​​of each dynamic parameter and substituting them into the comprehensive performance index of the aero-engine, a series of coupling performance and activity performance are obtained; by constructing a two-dimensional coordinate diagram of the coupling performance and activity performance, the area formed by the corresponding coordinate points of all coupling performance and activity performance in the coordinate diagram is the qualified interval of the comprehensive performance of the aero-engine.

8. An aircraft engine comprehensive performance evaluation device, comprising a processor, a memory, and a computer program stored in the memory; characterized in that: When the processor executes the computer program, it implements the method for comprehensive performance evaluation of aircraft engines based on test data and network resilience according to any one of claims 1 to 7.

9. A computer-readable storage medium, wherein a computer program is stored in the medium; characterized in that: When the computer program is executed by a processor, it implements the method for comprehensive performance evaluation of an aircraft engine based on test data and network resilience according to any one of claims 1 to 7.

Images