Aero-engine overall service life evaluation method based on multi-level service life data

Abstract

The invention provides an aero-engine overall service life evaluation method based on multi-level service life data, and relates to the field of evaluation of product service life distribution. Based on three-parameter Weibull life distribution, by adopting a sample reconstruction mode, from life data of key parts of all levels of an aero-engine, random sampling is conducted on the life data of parts of all the levels, and therefore the life data of a system is directly obtained. According to the method, propagation of errors of part reliability evaluation to system reliability evaluation is effectively blocked, so that the accuracy of system reliability evaluation is improved. A correlation model of position parameters and sample minimum values is further adopted, and a likelihood function model of scale parameters and a linear correlation model of shape parameters are combined, so that Weibull distribution parameters with higher precision are obtained. Compared with a traditional method, the reliability evaluation method under more accurate small sample data can be provided, and the reliability of an evaluation result is enhanced.

Description

Technical Field

[0001] The invention relates to the field of product life distribution evaluation, in particular to an aircraft engine whole machine life evaluation method based on multi-level life data. Background Art

[0002] In recent years, the reliability assessment of aircraft engines has received more and more attention, and the relevant reliability assessment methods are also known as one of the hot spots of current research. Aircraft engines are large and complex equipment, and the aircraft engine system consists of multiple subsystems. Each subsystem can be further divided into several parts, and these parts are composed of multiple components. This structure forms a multi-level aircraft engine system.

[0003] In view of the increase in high-reliability and long-life products, it is difficult for traditional high-level reliability tests to obtain complete failure data within a limited time. Especially for the whole engine, its high manufacturing cost limits the implementation of the whole engine reliability test. Therefore, studying how to use low-level life data to evaluate the reliability of high-level systems has become a research hotspot.

[0004] In the small sample reliability assessment method, due to the extremely small sample size, the traditional method is very likely to have the estimated value of the position parameter be 0. For this reason, reliability analysis methods for small sample data are proposed, which are mainly divided into two categories: one is to comprehensively utilize historical information and historical failure data of similar products, such as Bayesian method and semi-empirical method; the other is to use only current test data, mainly to improve the reliability assessment method under large sample size, so that it can be applied to the case of small samples, thereby improving the reliability assessment accuracy under small samples, such as Bootstrap method.

[0005] However, in the parameter estimation method of small sample life data, the accuracy of the semi-empirical method is limited by the expert knowledge and the quality of historical data; the Bayesian rule relies on the accuracy of prior information. If the prior information is insufficient or biased, the estimation results will be affected. In addition, although the sample expansion method improves the estimation accuracy by expanding the sample size, it may introduce duplicate samples in the resampling process because it is based on the original sample characteristics, especially in the case of small samples. This will amplify the deviation in the original data, thereby affecting the accuracy of the parameter estimation results.

[0006] In the current reliability assessment method of the system, the traditional series model calculates the reliability of the subsystem based on the reliability of the components, and then calculates the reliability of the whole system based on the reliability of the subsystem. Due to the small amount of failure data of each subsystem and component, the reliability assessment error is large. The use of a series connection method for system reliability assessment will cause the error in the component reliability estimation index to gradually expand through the product method, and ultimately affect the reliability assessment effect of the whole engine. For small sample conditions, this error will be further amplified.

[0007] Weibull distribution has been widely used in the reliability assessment of the life of various products such as aircraft engines, but it is usually based on the two-parameter Weibull distribution. However, for long-life products, the two-parameter Weibull distribution significantly underestimates the life and reliability of the product. Summary of the invention

[0008] In the prior art, when evaluating system reliability, errors propagate from components to the system and gradually expand, resulting in low accuracy of system reliability evaluation. In view of the shortcomings of the prior art, the present invention provides a more accurate aircraft engine whole life evaluation method based on multi-level life data, including:

[0009] Step 1: Determine a plurality of low-level key components of an aircraft engine system, wherein the low-level of the aircraft engine system is a level corresponding to components or assemblies of the aircraft engine system, and the key components are components or assemblies in the aircraft engine system that can affect system performance;

[0010] Step 2: For each key component, obtain multiple key components of the same model and conduct multiple tests to obtain multiple life data as a set of life data for the key component, and then obtain multiple sets of life data corresponding to multiple key components;

[0011] Step 3: randomly extract a life data from each of the multiple life data sets, take the minimum value of the multiple life data sets as the target data, and take the target data as the life data of the whole engine, the whole engine set corresponds to the high level of the whole aircraft engine system, repeat the above extraction operation to obtain n target data (t1, t2, ... t n ), where t1 is the target data obtained after the first extraction operation, t2 is the target data obtained after the second extraction operation, and t n is the target data obtained after the nth extraction operation;

[0012] Step 4: Based on the Weibull parameter estimation method with small samples based on the asymptotic approximation theory, for n target data (t1, t2, ... t n) to calculate and obtain the initial position parameters, initial scale parameters and initial shape parameters, and iteratively update the initial position parameters, initial scale parameters and initial shape parameters to obtain the final position parameters Final scale parameter and the final shape parameters That is, the Weibull parameter of the engine life distribution

[0013] Step 5: Weibull parameters based on the engine life distribution Obtaining the reliability formula Among them, t represents the time, is the reliability of the entire aircraft engine life at time t.

[0014] Optionally, step 4 specifically includes:

[0015] Step 4.1: Initialize the Weibull parameters to obtain the initial location parameter γ0, the initial scale parameter η0 and the initial shape parameter β0;

[0016] Step 4.2: Set the initial position parameter γ0 as the current position parameter γ, the initial scale parameter η0 as the current scale parameter η, the initial shape parameter β0 as the current shape parameter β, and set the initial number of iterations to 1;

[0017] Step 4.3: According to the current position parameter γ, current scale parameter η and current shape parameter β of the current iteration number, the deviation correction model of the position parameter, the likelihood function model of the scale parameter and the linear correlation model of the shape parameter are used to calculate the linear correlation model of the shape parameter based on the n target data (t1, t2, ..., t n ), update the current position parameter γ, the current scale parameter η and the current shape parameter β to obtain an updated position parameter γ', an updated shape parameter β' and an updated scale parameter η';

[0018] Step 4.4: Determine whether the difference between the updated position parameter γ' in the current iteration number and the position parameter of the previous iteration number of the current iteration number is less than or equal to the preset threshold ε. If the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is greater than the preset threshold ε, use the position parameter γ' as the current position parameter γ, the shape parameter β' as the current shape parameter β, and the scale parameter η' as the current scale parameter η, and return to execute step 4.3 until the difference between the updated position parameter γ' in the current iteration number and the updated position parameter in the previous iteration number is less than or equal to the preset threshold ε; if the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is less than or equal to the preset threshold ε, use the updated position parameter γ' as the final position parameter The updated shape parameter β' is used as the final shape parameter The updated scale parameter η' is used as the final scale parameter To obtain the Weibull parameters of the engine life distribution

[0019] Optionally, step 4.1 specifically includes:

[0020] The initial position parameter γ0 is set to 0, and the initial scale parameter η0 and the initial shape parameter β0 are calculated using the least squares method.

[0021] Optionally, the deviation correction model of the location parameter γ in step 4.3 is:

[0022] γ'=min(t i )-ηn -1 / β Γ(1+1 / β);

[0023] Among them, γ' is the updated position parameter, β is the current shape parameter, η is the current scale parameter, n is the number of target data, min(t i ) represents the minimum value among n target data.

[0024] Optionally, the likelihood function model of the scale parameter in step 4.3 is:

[0025]

[0026] Among them, η' is the updated scale parameter, β is the current shape parameter, n is the number of target data, γ is the current position parameter, t i is the i-th target data among n target data.

[0027] Optionally, the linear correlation model of the shape parameters in step 4.3 is:

[0028]

[0029] x i =ln(t i -γ);

[0030] y i =ln[-ln((1-F n (t i )))];

[0031]

[0032] Among them, β' is the updated shape parameter, n is the number of target data, t i is the i-th target data among the n target data, and γ is the current position parameter.

[0033] Optionally, the reliability formula in step 5 is:

[0034]

[0035] Among them, the Weibull parameter of the engine life distribution is Substitute the above formula to obtain the reliability formula of the entire engine.

[0036] The beneficial effects of adopting the above technical solution are:

[0037] 1. The present invention aims to improve the accuracy of aircraft engine system reliability assessment. To this end, the present invention is based on the three-parameter Weibull life distribution. By using sample reconstruction, starting from the perspective of aircraft engine key life data, the life data of components at each level is randomly sampled, thereby directly obtaining the life data of the system. This method effectively blocks the propagation of component reliability assessment errors to system reliability assessment, thereby improving the accuracy of system reliability assessment.

[0038] 2. The present invention adopts the correlation model between the location parameter and the minimum value of the sample, and combines the likelihood function model of the scale parameter and the linear correlation model of the shape parameter to obtain the Weibull distribution parameters with higher accuracy. Compared with the traditional method, the present invention can provide a more accurate reliability evaluation method under small sample data, and enhance the reliability of the evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flow chart of a method for evaluating the life of an aircraft engine based on multi-level life data in an embodiment of the present invention;

[0040] Figure 2A schematic diagram of the structure of an aircraft engine system in an embodiment of the present invention;

[0041] Figure 3 is a flow chart of an asymptotic approximation method in an embodiment of the present invention;

[0042] Figure 4 4 is a graph showing the changes of three Weibull parameters with the number of iterations in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0044] In view of the problems existing in the prior art, the present invention provides an aircraft engine whole machine life assessment method based on multi-level life data, combined with Figure 1 , which may include the following steps:

[0045] Step 1: Identify multiple low-level key components of the aircraft engine system.

[0046] Among them, the lower level of the aircraft engine system is the level corresponding to the parts or components of the engine, and the key parts are the key parts or components in the aircraft engine system that can affect the system performance. Specifically, the key parts are parts or components that are crucial to the function of the engine system. Failure of key parts will cause the performance of the entire system to decline. Identifying key parts is an important part of the FMECA process, which can improve the maintainability and safety of the system. Among them, the aircraft engine system is composed of subsystems such as the air inlet, combustion chamber, gas turbine, free turbine, compressor, and tail nozzle. According to FMECA analysis, these subsystems are considered to be the key parts of the engine system. In the specific implementation, combined with Figure 2 The engine body, that is, the entire engine system, can be understood as a high-level level, and the low-level level includes compressor, combustion chamber, tail nozzle, gas turbine, free turbine, transmission, axial flow compressor rotor blades, centrifugal compressor rotor blades, axial flow compressor rotor discs, centrifugal compressor rotor discs, tail nozzle, gas turbine second-stage rotor blades, gas turbine second-stage rotor discs, gas turbine first-stage rotor blades, gas turbine rear bearing, free turbine first-stage rotor disc, free turbine second-stage rotor disc, free turbine shaft, free turbine first-stage guide vane, free turbine first-stage rotor lock plate and output shaft.

[0047] It should be noted that all components listed in the above-mentioned low-level are key parts screened out for a certain aircraft engine, and the key parts determined by the present invention are multiple key parts therein.

[0048] Specifically, for the determination of key parts, the key parts constituting the entire aircraft engine can be determined from the above-mentioned low-level components or parts based on engineering experience, actual needs or FMECA analysis results.

[0049] Step 2: For each key component, obtain multiple key components of the same model and conduct multiple tests to obtain multiple life data as a set of life data for the key component, and then obtain multiple sets of life data corresponding to multiple key components;

[0050] If a certain engine is determined to have 5 key parts, 5 sets of life data can be obtained: and Among them, n1, n2, n3, n4 and n5 are the sample sizes of each key part, that is, the number of tests for each key part. For example, if the first key part among the five key parts is a gas turbine, then n1 gas turbines of the same model are tested n1 times to obtain a set of life data for the gas turbine.

[0051]

[0052] During specific implementation, the life data of key components in each level can also be determined based on reliability tests or failure data collected on-site.

[0053] Step 3: randomly extract a life data from each of the multiple life data sets, take the minimum value of the multiple life data sets as the target data, and take the target data as the life data of the whole engine, the whole engine set corresponds to the high level of the whole aircraft engine system, repeat the above extraction operation to obtain n target data (t1, t2, ... t n ), where t1 is the target data obtained after the first extraction operation, t2 is the target data obtained after the second extraction operation, and t n is the target data obtained after the nth extraction operation;

[0054] In the specific implementation, this step can be understood as the reconstruction of the sample of the whole engine. Specifically, according to the "principle of competition for unit life data at the same level", that is, the system life should be the minimum value of the unit life data of all key parts. Based on this idea, the Monte-Carlo sampling technology is used to randomly sample the life data of all key parts respectively, and the minimum value of each sampling result is selected to construct high-level life data, thereby realizing the sample reconstruction of the whole engine.

[0055] The life data of the high-level engine system is directly obtained by extracting it from the life data of the low-level key parts, so that the high-level system life data can be directly fitted, thereby avoiding the error propagation between the parameter fitting process from the key parts life distribution to the system life distribution.

[0056] Step 4: Based on the Weibull parameter estimation method with small samples based on the asymptotic approximation theory, for n target data (t1, t2, ... t n ) to calculate and obtain the initial position parameters, initial scale parameters and initial shape parameters, and iteratively update the initial position parameters, initial scale parameters and initial shape parameters to obtain the final position parameters Final scale parameter and the final shape parameters That is, the Weibull parameter of the engine life distribution

[0057] Combination Figure 3 , step 4 specifically includes the following steps:

[0058] Step 4.1: Initialize the Weibull parameters to obtain the initial location parameter γ0, the initial scale parameter η0 and the initial shape parameter β0;

[0059] Specifically, the initial position parameter γ0 is set to 0, and the initial scale parameter η0 and the initial shape parameter β0 are calculated using the least squares method.

[0060] Step 4.2: Set the initial position parameter γ0 as the current position parameter γ, the initial scale parameter η0 as the current scale parameter η, the initial shape parameter β0 as the current shape parameter β, and set the initial number of iterations to 1;

[0061] Step 4.3: According to the location parameter γ, scale parameter η and shape parameter β at the current iteration number, the deviation correction model of the location parameter, the likelihood function model of the scale parameter and the linear correlation model of the shape parameter are used to calculate the linear correlation model of the shape parameter based on the n target data (t1, t2, ..., t n ), update the current position parameter γ, the current scale parameter η and the current shape parameter β to obtain an updated position parameter γ', an updated shape parameter β' and an updated scale parameter η';

[0062] The deviation correction model of the location parameter γ is:

[0063] γ'=min(t i )-ηn -1 / β Γ(1+1 / β);

[0064] Among them, γ' is the updated position parameter, β is the current shape parameter, η is the current scale parameter, n is the number of target data, min(t i ) represents the minimum value among n target data.

[0065] The likelihood function model of the scale parameter is:

[0066]

[0067] Among them, η' is the updated scale parameter, β is the current shape parameter, n is the number of target data, γ is the current position parameter, t i is the i-th target data among n target data.

[0068] Optionally, the linear correlation model of the shape parameters in step 4.3 is:

[0069]

[0070] x i =ln(t i -γ);

[0071] y i =ln[-ln((1-F n (t i )))];

[0072]

[0073] Among them, β' is the updated shape parameter, n is the number of target data, t i is the i-th target data among the n target data, and γ is the current position parameter.

[0074] in, Figure 3 Medium j represents the position parameter when the number of iterations is j, η j represents the scale parameter when the number of iterations is j, β j represents the shape parameter when the number of iterations is j,

[0075] Step 4.4: Determine whether the difference between the updated position parameter γ' in the current iteration number and the position parameter of the previous iteration number of the current iteration number is less than or equal to the preset threshold ε. If the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is greater than the preset threshold ε, use the position parameter γ' as the current position parameter γ, the shape parameter β' as the current shape parameter β, and the scale parameter η' as the current scale parameter η, and return to execute step 4.3 until the difference between the updated position parameter γ' in the current iteration number and the updated position parameter in the previous iteration number is less than or equal to the preset threshold ε; if the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is less than or equal to the preset threshold ε, use the updated position parameter γ' as the final position parameter The updated shape parameter β' is used as the final shape parameter The updated scale parameter η' is used as the final scale parameter To obtain the Weibull parameters of the engine life distribution

[0076] in, Figure 3 Medium j-1 represents the position parameter when the number of iterations is j-1, Represents the estimated optimal value of the Weibull parameter.

[0077] Step 5: Weibull parameters based on the engine life distribution Obtaining the reliability formula Among them, t represents the time, is the reliability of the entire aircraft engine life at time t.

[0078] The reliability formula is:

[0079]

[0080] Among them, the Weibull parameter of the engine life distribution is Substitute the above formula to obtain the reliability formula of the entire engine.

[0081] Among them, steps 4 and 5 are equivalent to Figure 1 The reliability evaluation method under small samples is used to realize parameter estimation and reliability evaluation of Weibull distribution.

[0082] In a specific implementation, in order to evaluate the reliability of an aircraft engine using the failure data of key parts, it is assumed that the level is 2, the number of key parts in the level is 5, namely key parts ABCDE, and the amount of failure data of each key part is 6. As shown in Table 1 below. The purpose of the present invention is to use the failure data of these five key parts to evaluate the reliability of a system composed of these five key parts.

[0083] First, the number of levels is determined to be 2, and the number of low-level key parts is 5, and then 5 sets of life data of 5 key parts are determined. Table 1 shows the life data of 5 key parts;

[0084] Table 1 Life data of 5 key parts

[0085]

[0086]

[0087] Then, the system sample is reconstructed. For the life data of the five key parts in Table 1, the samples are randomly sampled for the first time, and the samples are: 1.619, 1.189, 2.421, 2.841, 4.565; the minimum value is 1.189 for the system sample; the samples are randomly sampled for the second time, and the samples are: 2.410, 1.892, 2.424, 2.506, 2.224; the minimum value is 1.892 for the system sample; after six samplings, the system samples are: 1.189, 1.892, 1.940, 1.619, 2.251, 1.921;

[0088] After that, the initial values ​​of the Weibull parameters are calculated first. The location parameter is 0, the scale parameter is 1.9708, and the shape parameter is 4.6159. Then, the Weibull parameters are calculated iteratively using the calculation formulas of the three parameters. The convergence process is as follows: Figure 4 As shown. The final Weibull parameters are (0.410, 1.473, 3.383). The reliability calculated at 1.4 is 0.770;

[0089] For the existing methods, the Weibull parameters obtained by asymptotic approximation are (1.082, 1.123, 1.576), (0.621, 2.969, 1.109), (1.55, 0.954, 2.314), (1.702, 1.21, 2.306), and (0.637, 3.669, 2.493), respectively. The reliabilities calculated at 1.4 are 0.872, 0.797, 0.986, 0.960, and 0.980, respectively, and the corresponding system reliability is 0.645;

[0090] These data are derived from the Weibull parameter true values ​​(1,2,2), (1,3,2), (1,2,3), (1,2,4) and (1,4,2), and the corresponding reliability true values ​​at 1.4 are 0.961, 0.982, 0.992, 0.998 and 0.990, respectively, and the corresponding system reliability is 0.925.

[0091] By comparing the system reliability calculated with the true value, it can be found that the reliability of 0.770 calculated by the aircraft engine whole life assessment method based on multi-level life data is closest to the true value system reliability of 0.925, which is 55.36% higher than the accuracy of the existing method.

[0092] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) to form a technical solution.

Claims

1. A method for evaluating the life of an aircraft engine based on multi-level life data, characterized in that: include: Step 1: Determine a plurality of low-level key components of an aircraft engine system, wherein the low-level of the aircraft engine system is a level corresponding to components or assemblies of the aircraft engine system, and the key components are components or assemblies in the aircraft engine system that can affect system performance; Step 2: For each key component, obtain multiple key components of the same model and conduct multiple tests to obtain multiple life data as a set of life data for the key component, and then obtain multiple sets of life data corresponding to multiple key components; Step 3: randomly extract a life data from each of the multiple life data sets, take the minimum value of the multiple life data sets as the target data, and take the target data as the life data of the whole engine, the whole engine set corresponds to the high level of the whole aircraft engine system, repeat the above extraction operation to obtain n target data (t1, t2, ... t n ), where t1 is the target data obtained after the first extraction operation, t2 is the target data obtained after the second extraction operation, and t n is the target data obtained after the nth extraction operation; Step 4: Based on the Weibull parameter estimation method with small samples based on the asymptotic approximation theory, for n target data (t1, t2, ... t n ) to calculate and obtain the initial position parameters, initial scale parameters and initial shape parameters, and iteratively update the initial position parameters, initial scale parameters and initial shape parameters to obtain the final position parameters Final scale parameter And the final shape parameters That is, the Weibull parameter of the engine life distribution Step 5: Weibull parameters based on the engine life distribution Obtaining the reliability formula Among them, t represents the time, is the reliability of the entire aircraft engine life at time t.

2. The aircraft engine whole machine life assessment method based on multi-level life data according to claim 1 is characterized in that: Step 4 specifically includes: Step 4.1: Initialize the Weibull parameters to obtain the initial location parameter γ0, the initial scale parameter η0 and the initial shape parameter β0; Step 4.2: Set the initial position parameter γ0 as the current position parameter γ, the initial scale parameter η0 as the current scale parameter η, the initial shape parameter β0 as the current shape parameter β, and set the initial number of iterations to 1; Step 4.3: According to the current position parameter γ, current scale parameter η and current shape parameter β of the current iteration number, the deviation correction model of the position parameter, the likelihood function model of the scale parameter and the linear correlation model of the shape parameter are used to calculate the linear correlation model of the shape parameter based on the n target data (t1, t2, ..., t n ), update the current position parameter γ, the current scale parameter η and the current shape parameter β to obtain an updated position parameter γ', an updated shape parameter β' and an updated scale parameter η'; Step 4.4: Determine whether the difference between the updated position parameter γ' in the current iteration number and the position parameter of the previous iteration number of the current iteration number is less than or equal to the preset threshold ε. If the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is greater than the preset threshold ε, use the position parameter γ' as the current position parameter γ, the shape parameter β' as the current shape parameter β, and the scale parameter η' as the current scale parameter η, and return to execute step 4.3 until the difference between the updated position parameter γ' in the current iteration number and the updated position parameter in the previous iteration number is less than or equal to the preset threshold ε; if the difference between the updated position parameter γ' in the current iteration number and the position parameter in the previous iteration number of the current iteration number is less than or equal to the preset threshold ε, use the updated position parameter γ' as the final position parameter The updated shape parameter β' is used as the final shape parameter The updated scale parameter η' is used as the final scale parameter To obtain the Weibull parameters of the engine life distribution 3. The method for evaluating the life of an aircraft engine based on multi-level life data according to claim 2, characterized in that: Step 4.1 specifically includes: The initial position parameter γ0 is set to 0, and the initial scale parameter η0 and the initial shape parameter β0 are calculated using the least squares method.

4. The method for evaluating the life of an aircraft engine based on multi-level life data according to claim 2, characterized in that: The deviation correction model of the location parameter γ is: γ'=min(t i )-n -1 / β C(1+1 / b) Among them, γ' is the updated position parameter, β is the current shape parameter, η is the current scale parameter, n is the number of target data, min(t i ) represents the minimum value among n target data.

5. The method for evaluating the life of an aircraft engine based on multi-level life data according to claim 2, characterized in that: The likelihood function model of the scale parameter is: Among them, η' is the updated scale parameter, β is the current shape parameter, n is the number of target data, γ is the current position parameter, t i is the i-th target data among n target data.

6. The method for evaluating the life of an aircraft engine based on multi-level life data according to claim 2, characterized in that: The linear correlation model of the shape parameters is: x i =ln(t i -c); y i =ln[-ln((1-F n (t i )))]; Among them, β' is the updated shape parameter, n is the number of target data, t i is the i-th target data among the n target data, and γ is the current position parameter.

7. The method for evaluating the life of an aircraft engine based on multi-level life data according to claim 1, characterized in that: The reliability formula in step 5 is: Among them, the Weibull parameter of the engine life distribution is Substitute the above formula to obtain the reliability formula of the entire engine.

Images