A method and device for updating turbine blade failure probability based on output observation data

By using a method based on the Dirac function to update the failure probability of the turbine blade structure with the output observation data, the problem of the lack of an effective update method in the prior art is solved, and accurate failure risk assessment is achieved under the condition of scarce early observation data, thereby improving computational efficiency and applicability.

CN123116507APending Publication Date: 2026-11-17NORTHWESTERN POLYTECHNICAL UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611617579.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-19
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies lack effective methods for updating the failure probability of turbine blade structures based on output observation data, especially in the early stages of turbine blade structure service when observation data is scarce, making it difficult to accurately assess its failure risk.

Method used

A Dirac function-based approach is adopted to construct the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure by outputting observation data. The integral expression of the output response condition density function is derived by utilizing the screening property of the Dirac function. A single reliability analysis is performed using the Dirac function normal approximation technique to calculate the estimated value of the output response condition density function. Finally, the failure probability is calculated using the Bayesian update method.

Benefits of technology

It enables efficient updating of the failure probability of turbine blade structures under the condition of output observation data, accurately characterizes its true reliability level, reduces the time and economic cost of repeated reliability analysis, and improves the computational efficiency and engineering applicability of failure risk assessment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123116507A_ABST
    Figure CN123116507A_ABST
Patent Text Reader

Abstract

This invention discloses a method and device for updating the failure probability of turbine blades based on output observation data, belonging to the field of reliability analysis technology. It specifically updates the failure probability based on output observation data, derives the integral expression of the output response conditional density function using the screening property of the Dirac delta function, and employs the normal approximation method of the Dirac delta function to achieve efficient estimation of the output response conditional density function value. Only a single reliability analysis is needed to obtain all the output response conditional density function values ​​required to construct the likelihood function of the uncertainty distribution parameters, thereby completing the solution of the likelihood function of the uncertainty distribution parameters and the Bayesian update of the turbine blade structural failure probability. This invention not only fully utilizes readily available output observation data in engineering and closely matches the actual service conditions of turbine blades, but also overcomes the key challenge of efficiently solving the output response density function, significantly reducing the time and economic costs associated with repeated reliability analyses.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of reliability analysis technology, specifically relating to a method and device for updating the failure probability of turbine blades based on output observation data. Background Technology

[0002] Currently, when acquiring new observational data, the probability distribution of random input changes, necessitating updates to the failure probability of the turbine blade structure to accurately assess its true reliability level in real time. This is especially crucial during the early stages of turbine blade operation when observational data is scarce; efficient failure probability update methods are essential for accurately predicting turbine blade structure failure risks.

[0003] However, existing failure probability update methods are still insufficient. Most existing failure probability update methods are adapted to newly added input observation data, while in actual engineering, it is more common to add new output observation data. For the failure probability update problem in this case, the existing technology still lacks an effective failure probability update method based on output observation data for turbine blade structures. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for updating the failure probability of turbine blades based on output observation data, so as to solve the problem that there is no effective method for updating the failure probability of turbine blade structures based on output observation data in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: The function to obtain the turbine blade structure, output the observation dataset, the probability density function of random input under given uncertainty distribution parameters, and the prior uniform distribution probability density function of the uncertainty distribution parameters; Based on the probability density function and the prior uniform distribution probability density function, a joint density function of the random input and uncertainty distribution parameters of the turbine blade structure is constructed. Based on the joint density function and the function of function, the Dirac delta is introduced. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. A joint distribution sample set of the joint density function is extracted using combined sampling techniques, and the corresponding function value is calculated. The Dirac delta method is employed. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; Based on the estimated values ​​of the conditional density function of the output response at all output observation data values ​​and the output observation dataset, calculate the likelihood function of the uncertainty distribution parameters; Based on the function, the joint density function, and the likelihood function, a failure probability update formula for the turbine blade structure is constructed, and the updated failure probability of the turbine blade structure is calculated according to the failure probability update formula.

[0006] Furthermore, the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure is:

[0007] in, Represents a random input vector and the parameter vector of the uncertainty distribution The joint density function, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, For random input vectors The realized value, The parameter vector of the uncertainty distribution The realized value.

[0008] Furthermore, the Dirac is introduced. The function derivation includes the integral calculation of the output response conditional density function at each output observation value, including: using Dirac... The filtering property of the function transforms the conditional density function of the output response at each output observation value into the following integral formula:

[0009] in, Indicates a given one-dimensional random output as well as The conditional density function of the output response at that time. It outputs the observation dataset. The output observation data values ​​in the middle, One-dimensional random output The realized value, The parameter vector of the uncertainty distribution The realized value, subscript It outputs the observation dataset. Data value index, superscript It outputs the observation dataset. The label, It outputs the observation dataset. The capacity, superscript It is the transpose symbol. For the Dirac function, Indicates a function. For random input vectors The realized value, For the parameter vector corresponding to the uncertainty distribution The integral variable, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, Represents a random input vector The integration domain, Represents the parameter vector of the uncertainty distribution The integration domain, These are the normalized coefficients of the distribution parameters. .

[0010] Furthermore, the combined sampling technique is used to extract a joint distribution sample set of the joint density function and calculate the corresponding function value, including: The Monte Carlo sample set of the uncertainty distribution parameter vector is extracted from the prior uniform distribution probability density function; For each sample of the extracted uncertainty distribution parameter vector, a sample of the random input vector is randomly drawn from the probability density function of the random input vector corresponding to the sample of the uncertainty distribution parameter vector. This results in a joint distribution sample set of the random input vector and the uncertainty distribution parameter vector, and the function value corresponding to each sample in the joint distribution sample set is calculated.

[0011] Furthermore, the Dirac method is adopted. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis, including: The Dirac function in the integral formula and The output response conditional density function is approximated using the normal distribution probability density function for the random input vector and the normal distribution probability density function for the uncertain distribution parameter vector, respectively. Using the joint distribution sample set and the corresponding function values, the estimated value of the output response conditional density function is calculated using the following formula:

[0012] in, Represents the conditional density function of the output response The estimated value, superscript For the joint distribution sample set The label, The parameter vector of the uncertainty distribution The realized value, for The One portion, for medium sample The Each component, subscript It is a jointly distributed sample set The sample index in The parameter vector of the uncertainty distribution dimensionality yes medium sample The corresponding function value, It is a jointly distributed sample set The sample size Therefore Let be the normal distribution probability density function with smoothing parameters. Therefore Let be the normal distribution probability density function with smoothing parameters. The normalization coefficient constant of the distribution parameters The estimated value, ; The single reliability analysis refers to using the same set of jointly distributed sample sets. and the corresponding function values , All output observation data values ​​can be obtained at once. The corresponding output response conditional density function The estimated value The superscript T is the transpose symbol.

[0013] Furthermore, based on the estimated values ​​of the output response conditional density function at all output observation data values ​​and the output observation dataset, the likelihood function of the uncertainty distribution parameter is calculated, including: using the estimated values ​​of the output response conditional density function, the likelihood function of the uncertainty distribution parameter is obtained by multiplication.

[0014] in, Indicates based on the output observation dataset Regarding the parameter vector of the uncertainty distribution likelihood function The estimated value.

[0015] Furthermore, based on the function, the joint density function, and the likelihood function, a failure probability update formula for the turbine blade structure is constructed. The updated failure probability of the turbine blade structure is calculated using this formula, including: A Bayesian update transform is performed on the failure probability of the turbine blade structure to obtain the integral expression of the updated failure probability:

[0016] in, Indicates based on the output observation dataset Failure probability after turbine blade structure update, subscript Represents function The failure domain, superscript Indicates updating the label. It is a failure domain The indicator function, if ,but ,otherwise , The normalization constant is , Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function; Using the jointly distributed sample set, an estimated value for the failure probability of the updated turbine blade structure is obtained:

[0017] in, for The estimated value, yes exist The function value at that point, It is a failure domain Indicator function exist The function value at that point, if ,but ,otherwise , Normalization constant The estimated value, .

[0018] Secondly, the present invention provides a turbine blade failure probability update system based on output observation data, comprising an acquisition module, a construction module, a first calculation module, a second calculation module, and an update module, wherein: Acquisition module: used to acquire the function of turbine blade structure, output observation dataset, probability density function of random input under given uncertainty distribution parameters, and prior uniform distribution probability density function of uncertainty distribution parameters; Construction module: used to construct the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure based on the probability density function and the prior uniform distribution probability density function; First calculation module: used to introduce Dirac based on the joint density function and the function function. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. A joint distribution sample set of the joint density function is extracted using combined sampling techniques, and the corresponding function value is calculated. The Dirac delta method is employed. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; The second calculation module is used to calculate the likelihood function of the uncertainty distribution parameter based on the estimated value of the conditional density function of the output response at all output observation data values ​​and the output observation dataset. The update module is used to construct a failure probability update formula for the turbine blade structure based on the function, the joint density function, and the likelihood function, and to calculate the failure probability of the updated turbine blade structure according to the failure probability update formula.

[0019] Thirdly, a terminal device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0020] Fourthly, a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method.

[0021] Compared with the prior art, the present invention has the following beneficial technical effects: This invention provides a method for updating the failure probability of turbine blades based on output observation data. Most current mainstream failure probability update methods rely on input observation data for extrapolation. Although the density function of randomly input data is known and computationally convenient, it is easier to collect output observation data such as stress and deformation of turbine blades in engineering settings. However, it is difficult to obtain the underlying measured information of materials and loads. Traditional methods suffer from insufficient adaptability to engineering scenarios. Furthermore, directly using output observation data to update the failure probability faces the core technical bottleneck of an unknown output response density function, making efficient solution difficult. This invention specifically uses output observation data as the basis for failure probability updating, utilizing the Dirac delta method... The sieving property of the function yields the integral expression of the conditional density function of the output response, and the Dirac delta method is used. The normal approximation method achieves efficient estimation of the output response conditional density function value. Only a single reliability analysis is needed to obtain all the output response conditional density function values ​​required to construct the likelihood function of the uncertainty distribution parameters. This allows for the solution of the likelihood function of the uncertainty distribution parameters and the Bayesian update of the turbine blade structure failure probability. This invention fully utilizes readily available output observation data in engineering, closely reflects the actual service conditions of turbine blades, and overcomes the key challenge of efficiently solving the output response density function. It significantly reduces the time and economic costs associated with repeated reliability analyses. The probability distribution of random inputs can be dynamically corrected as observation data accumulates, resulting in a more realistic updated failure probability for the turbine blade structure. This accurately characterizes the true reliability level of the turbine blade. Even under conditions of scarce observation data in the early stages of service, it can output stable and reliable calculation results, effectively improving the computational efficiency and engineering applicability of turbine blade structure failure risk assessment. Attached Figure Description

[0022] Figure 1 This is a flowchart of a turbine blade structure failure probability update method based on output observation data in an embodiment of the present invention; Figure 2 This is a schematic diagram of a turbine blade structure according to an embodiment of the present invention; Figure 3 This is a sub-flowchart of a turbine blade structure failure probability update method based on output observation data in an embodiment of the present invention; Figure 4 This is another sub-flowchart of a turbine blade structure failure probability update method based on output observation data in an embodiment of the present invention; Figure 5 A schematic diagram is shown illustrating the method for calculating the updated failure probability of a turbine blade structure based on the output observation data according to this exemplary embodiment. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0025] It should be noted that the terms "first" and "second" in the specification and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0026] The present invention will now be described in further detail with reference to the accompanying drawings: like Figure 1 As shown, the turbine blade structure failure probability update method based on output observation data includes the following steps: Step 1: Obtain the function of the turbine blade structure, the output observation dataset, the probability density function of the random input under given uncertainty distribution parameters, and the prior uniform distribution probability density function of the uncertainty distribution parameters. Step 2: Construct the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure based on the probability density function and the prior uniform distribution probability density function; Turbine blades are critical components in aero-engines, gas turbines, and other similar equipment. Mounted on a rotor, they convert the kinetic and thermal energy of high-temperature, high-pressure gas into mechanical energy through high-speed rotation, thereby driving the turbine. Various uncertainties arise during the analysis, simulation, or optimization of turbine blade structures. These uncertainties include material properties, geometric dimensions, and operating conditions. Furthermore, some distribution parameters of the random inputs may also be uncertain. This exemplary embodiment first constructs the joint density function of the random inputs and uncertain distribution parameters of the turbine blade structure, and then uses this joint density function to update the failure probability calculation.

[0027] Turbine blade structure such as Figure 2 As shown in this exemplary embodiment, the number of random inputs is seven, specifically representing the yield strength. Young's modulus The coefficient of thermal expansion of nickel-based alloys Poisson's ratio Blade pressure side load Blade intake side load and thermal conductivity It should be noted that this exemplary embodiment is for illustrative purposes only, and in practical applications, the number of random inputs is not limited to seven.

[0028] In this exemplary embodiment, Table 1 shows the distribution form and distribution parameters of each random input: Table 1. Distribution form and distribution parameters of each random input

[0029] Table 1 shows the pressure side load on the blade. and blade intake side load The mean values ​​of all are parameters of an uncertain distribution, and their prior distributions are all uniform distributions.

[0030] Step 3: Based on the joint density function and the function of function, introduce the Dirac delta method. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. It then uses combined sampling techniques to extract a joint distribution sample set of the joint density function and calculates the corresponding function value, employing the Dirac delta method. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; The function function of the turbine blade structure refers to the function used to calculate the failure probability of the turbine blade structure. For example, based on the random input of the turbine blade structure shown in Table 1, the function function of the turbine blade structure can be expressed as:

[0031] in, It's about random input ( The implicit function of is determined by finite element analysis.

[0032] The joint density function of the random input and uncertainty distribution parameters of the turbine blade structure in this exemplary embodiment, as well as the function function, can be used not only to calculate the conditional density function of the output response of the turbine blade structure, but also to calculate the initial failure probability of the turbine blade structure.

[0033] Specifically, in an exemplary embodiment, such as Figure 3 As shown, step three specifically includes the following steps: Step S310, introduce Dirac The integral formula for calculating the conditional density function of the output response at each output observation data value is derived. Step S320: Use combined sampling techniques to extract a joint distribution sample set of the joint density function and calculate the corresponding function value; Step S330, using Dirac The normal approximation technique for the function, based on a joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis.

[0034] In an exemplary embodiment, the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure for:

[0035] in, Represents a random input vector and the parameter vector of the uncertainty distribution The joint probability density function, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, For random input vectors The realized value, The parameter vector of the uncertainty distribution The realized value.

[0036] In step S310, using Dirac The filtering property of the function selects the output response conditional density function at each output observation value. Transformed into the following integral formula:

[0037] in, Indicates a given one-dimensional random output as well as The conditional density function of the output response at that time. It outputs the observation dataset. The output observation data values ​​in the middle, One-dimensional random output The realized value, subscript It outputs the observation dataset. Data value index, superscript It outputs the observation dataset. The label, It outputs the observation dataset. The capacity, superscript It is the transpose symbol. For the Dirac function, Indicates a function. For the parameter vector corresponding to the uncertainty distribution The integral variable, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, Represents a random input vector The integration domain, Represents the parameter vector of the uncertainty distribution The integration domain, These are the normalized coefficients of the distribution parameters. .

[0038] In step S320, a Monte Carlo sample set of the uncertainty distribution parameter vector is extracted from the prior uniform distribution probability density function; For each sample of the extracted uncertainty distribution parameter vector, a sample of the random input vector is randomly drawn from the probability density function of the random input vector corresponding to the sample of the uncertainty distribution parameter vector. This results in a joint distribution sample set of the random input vector and the uncertainty distribution parameter vector, and the function value corresponding to each sample in the joint distribution sample set is calculated.

[0039] Specifically, firstly, through the uncertainty distribution parameter vector The prior uniform probability density function Extract Monte Carlo sample set Then for each extracted samples , due to The corresponding random input vector when the condition is met. probability density function To randomly select A sample Thus obtain and joint density function Joint distribution sample set And calculate the joint distribution sample set. random input vector samples Corresponding function value subscript For the jointly distributed sample set Index of the middle sample, superscript For the jointly distributed sample set The label, It is a jointly distributed sample set The sample size The value of depends on the order of magnitude of the failure probability of the turbine blade structure, and the superscript . It is the transpose symbol.

[0040] In step S330, the Dirac function in equation (3) above is... and The output response conditional density function is approximated using the normal distribution probability density function for the random input vector and the normal distribution probability density function for the uncertain distribution parameter vector, respectively. Using the joint distribution sample set and the corresponding function values, the estimated value of the output response conditional density function is calculated using the following formula:

[0041] in, Represents the conditional density function of the output response The estimated value, The parameter vector of the uncertainty distribution The realized value, for The One portion, for medium sample The One portion, The parameter vector of the uncertainty distribution dimensionality yes medium sample The corresponding function value, Therefore Let be the normal distribution probability density function with smoothing parameters. Therefore Let be the normal distribution probability density function with smoothing parameters. Normalized coefficients of the distribution parameters The estimated value, ; A single reliability analysis refers to using the same set of jointly distributed sample sets. and the corresponding function values To simultaneously obtain all output observation data values Corresponding output response conditional density function The estimated value .

[0042] In this exemplary embodiment, due to Dirac The function has a poor functional form, the Dirac function. and We need to use the normal distribution probability density function for the random input vector and the normal distribution probability density function for the uncertain distribution parameter vector to approximate the estimation.

[0043] Specifically, the Dirac function and They can be represented as:

[0044]

[0045] in, and All are smoothing constants. and Based on experience, these are all taken as relatively small constants; in this embodiment, they are taken as 0.05. The parameter vector of the uncertainty distribution Integral variable The Each component.

[0046] Step 4: Based on the estimated values ​​of the output response conditional density function at all output observation data points and the output observation dataset, calculate the likelihood function of the uncertainty distribution parameters: In one exemplary embodiment, the estimated value of the output response conditional density function is utilized. The uncertainty distribution parameters are obtained by multiplying together. The estimated value of the likelihood function:

[0047] in, Indicates based on the output observation dataset Regarding the parameter vector of the uncertainty distribution likelihood function The estimated value.

[0048] Step 5: Based on the function, joint density function, and likelihood function, construct the failure probability update formula for the turbine blade structure, and calculate the updated failure probability of the turbine blade structure according to the failure probability update formula.

[0049] Specifically, in an exemplary embodiment, such as Figure 4 As shown, step five specifically includes the following steps: Step S510: Perform a Bayesian update transformation on the failure probability of the turbine blade structure to obtain the integral expression of the updated failure probability. Step S520: Using the joint distribution sample set, obtain an estimate of the failure probability after the turbine blade structure is updated.

[0050] In step S510, a Bayesian update transformation is performed on the failure probability of the turbine blade structure to obtain the updated failure probability of the turbine blade structure. The integral expression is:

[0051] in, Indicates based on the output observation dataset Failure probability after turbine blade structure update, subscript Represents function The failure domain, superscript Indicates updating the label. It is a failure domain The indicator function, if ,but ,otherwise , Indicates based on the output observation dataset Uncertainty distribution parameter vector The posterior probability density function, The normalization constant is ; In step S520, the joint distribution sample set is used. and the likelihood function estimate of the uncertainty distribution parameters The estimated value for the failure probability after the turbine blade structure update is as follows:

[0052] in, for The estimated value, yes exist The function value at that point, It is a failure domain Indicator function exist The function value at that point, if ,but ,otherwise , Normalization constant The estimated value, .

[0053] Figure 5 A schematic diagram is shown illustrating the method for calculating the updated failure probability of a turbine blade structure based on output observation data, according to this exemplary embodiment. The results show that, with the accumulation of output observation data, the updated failure probability of the turbine blade structure is significantly greater than the initial failure probability, thus more accurately characterizing the true reliability level of the current turbine blade structure.

[0054] Based on the above description, in this exemplary embodiment, the function function of the turbine blade structure, the output observation dataset, the probability density function of the random input under given uncertainty distribution parameters, and the prior uniform distribution probability density function of the uncertainty distribution parameters are obtained; a joint density function of the random input and uncertainty distribution parameters of the turbine blade structure is constructed based on the probability density function and the prior uniform distribution probability density function; and a Dirac delta function is introduced based on the joint density function and the function function. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. It then uses combined sampling techniques to extract a joint distribution sample set of the joint density function and calculates the corresponding function value, employing the Dirac delta method. This exemplary embodiment proposes an effective method for updating the failure probability of turbine blade structures based on output observation data. This method uses a normal approximation technique, based on a joint distribution sample set and the corresponding function function values, to calculate the estimated output response conditional density function at all output observation data values ​​through a single reliability analysis. Based on the estimated output response conditional density function at all output observation data values ​​and the output observation dataset, the likelihood function of the uncertainty distribution parameters is calculated. A failure probability update formula for the turbine blade structure is constructed based on the function function, joint density function, and likelihood function. The updated failure probability of the turbine blade structure is then calculated using this formula. Specifically, this exemplary embodiment provides an effective method for updating the failure probability of turbine blade structures based on output observation data. This provides a reference and guidance for real-time assessment of the true reliability level of turbine blade structures, thereby saving significant economic and time costs. Specifically, the failure probability update method based on output observation data provided in this exemplary embodiment can calculate the estimated output response conditional density function at all output observation data values ​​used to construct the likelihood function of the uncertainty distribution parameters through a single reliability analysis, and then solve for the updated failure probability of the turbine blade structure. This solution method can effectively and quickly calculate the updated failure probability result of the turbine blade structure based on output observation data.

[0055] In another embodiment of the present invention, a turbine blade failure probability update system based on output observation data is provided, comprising an acquisition module, a construction module, a first calculation module, a second calculation module, and an update module, wherein: Acquisition module: used to acquire the function of turbine blade structure, output observation dataset, probability density function of random input under given uncertainty distribution parameters, and prior uniform distribution probability density function of uncertainty distribution parameters; Module: Used to construct the joint density function of random input and uncertainty distribution parameters of turbine blade structure based on probability density function and prior uniform distribution probability density function; First calculation module: used to introduce Dirac based on joint density function and function. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. A joint distribution sample set of the joint density function is extracted using combined sampling techniques, and the corresponding function value is calculated. The Dirac delta method is employed. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; The second calculation module is used to calculate the likelihood function of the uncertainty distribution parameters based on the estimated value of the conditional density function of the output response at all output observation data values ​​and the output observation dataset. The update module is used to construct a failure probability update formula for the turbine blade structure based on the function, joint density function, and likelihood function, and to calculate the failure probability of the turbine blade structure after the update according to the failure probability update formula.

[0056] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can be implemented in one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs) containing computer-usable program code. The form of a computer program product implemented on ROM (optical memory, etc.).

[0057] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that after reading the present invention, they can still make various changes, modifications or equivalent substitutions to the specific implementation of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the present invention pending approval.

Claims

1. A method for updating the failure probability of turbine blades based on output observation data, characterized in that, Includes the following steps: The function to obtain the turbine blade structure, output the observation dataset, the probability density function of random input under given uncertainty distribution parameters, and the prior uniform distribution probability density function of the uncertainty distribution parameters; Based on the probability density function and the prior uniform distribution probability density function, a joint density function of the random input and uncertainty distribution parameters of the turbine blade structure is constructed. Based on the joint density function and the function of function, the Dirac delta is introduced. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. A joint distribution sample set of the joint density function is extracted using combined sampling techniques, and the corresponding function value is calculated. The Dirac delta method is employed. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; Based on the estimated values ​​of the conditional density function of the output response at all output observation data values ​​and the output observation dataset, calculate the likelihood function of the uncertainty distribution parameters; Based on the function, the joint density function, and the likelihood function, a failure probability update formula for the turbine blade structure is constructed, and the updated failure probability of the turbine blade structure is calculated according to the failure probability update formula.

2. The turbine blade failure probability update method based on output observation data according to claim 1, characterized in that, The joint density function of the random input and uncertainty distribution parameters of the turbine blade structure is: in, Represents a random input vector and the parameter vector of the uncertainty distribution The joint density function, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, For random input vectors The realized value, The parameter vector of the uncertainty distribution The realized value.

3. The turbine blade failure probability update method based on output observation data according to claim 1, characterized in that, Introducing the Dirac The function derivation includes the integral calculation of the output response conditional density function at each output observation value, including: using Dirac... The filtering property of the function transforms the conditional density function of the output response at each output observation value into the following integral formula: in, Indicates a given one-dimensional random output as well as The conditional density function of the output response at that time. It outputs the observation dataset. The output observation data values ​​in the middle, One-dimensional random output The realized value, The parameter vector of the uncertainty distribution The realized value, subscript It outputs the observation dataset. Data value index, superscript It outputs the observation dataset. The label, It outputs the observation dataset. The capacity, superscript It is the transpose symbol. For the Dirac function, Indicates a function. For random input vectors The realized value, For the parameter vector corresponding to the uncertainty distribution The integral variable, Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function, Represents a random input vector The integration domain, Represents the parameter vector of the uncertainty distribution The integration domain, These are the normalized coefficients of the distribution parameters. .

4. The turbine blade failure probability update method based on output observation data according to claim 1, characterized in that, The method of combined sampling is used to extract a joint distribution sample set of the joint density function and calculate the corresponding function value, including: The Monte Carlo sample set of the uncertainty distribution parameter vector is extracted from the prior uniform distribution probability density function; For each sample of the extracted uncertainty distribution parameter vector, a sample of the random input vector is randomly drawn from the probability density function of the random input vector corresponding to the sample of the uncertainty distribution parameter vector. This results in a joint distribution sample set of the random input vector and the uncertainty distribution parameter vector, and the function value corresponding to each sample in the joint distribution sample set is calculated.

5. The turbine blade failure probability update method based on output observation data according to claim 3, characterized in that, Using the Dirac The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis, including: The Dirac function in the integral formula and The output response conditional density function is approximated using the normal distribution probability density function for the random input vector and the normal distribution probability density function for the uncertain distribution parameter vector, respectively. Using the joint distribution sample set and the corresponding function values, the estimated value of the output response conditional density function is calculated using the following formula: in, Represents the conditional density function of the output response The estimated value, superscript For the joint distribution sample set The label, The parameter vector of the uncertainty distribution The realized value, for The One portion, for medium sample The Each component, subscript It is a jointly distributed sample set The sample index in The parameter vector of the uncertainty distribution dimensionality yes medium sample The corresponding function value, It is a jointly distributed sample set The sample size Therefore Let be the normal distribution probability density function with smoothing parameters. Therefore Let be the normal distribution probability density function with smoothing parameters. Normalized coefficients of the distribution parameters The estimated value, ; The single reliability analysis refers to using the same set of jointly distributed sample sets. and the corresponding function values , All output observation data values ​​can be obtained at once. The corresponding output response conditional density function The estimated value The superscript T is the transpose symbol.

6. The turbine blade failure probability update method based on output observation data according to claim 5, characterized in that, Based on the estimated values ​​of the output response conditional density function at all output observation data values ​​and the output observation dataset, the likelihood function of the uncertainty distribution parameter is calculated, including: using the estimated values ​​of the output response conditional density function, the likelihood function of the uncertainty distribution parameter is obtained by multiplication. in, Indicates based on the output observation dataset Regarding the parameter vector of the uncertainty distribution likelihood function The estimated value.

7. The turbine blade failure probability update method based on output observation data according to claim 6, characterized in that, Based on the aforementioned function, joint density function, and likelihood function, a failure probability update formula for the turbine blade structure is constructed. The updated failure probability of the turbine blade structure is then calculated using this formula, including: A Bayesian update transform is performed on the failure probability of the turbine blade structure to obtain the integral expression of the updated failure probability: in, Indicates based on the output observation dataset Failure probability after turbine blade structure update, subscript Represents function The failure domain, superscript Indicates updating the label. It is a failure domain The indicator function, if ,but ,otherwise , The normalization constant is , Represents a random input vector Given an uncertainty distribution parameter vector The probability density function under the given conditions Represents the parameter vector of the uncertainty distribution The prior uniform probability density function; Using the jointly distributed sample set, an estimated value for the failure probability of the updated turbine blade structure is obtained: in, for The estimated value, yes exist The function value at that point, It is a failure domain Indicator function exist The function value at that point, if ,but ,otherwise , Normalization constant The estimated value, .

8. A turbine blade failure probability update system based on output observation data, characterized in that, It includes an acquisition module, a construction module, a first calculation module, a second calculation module, and an update module, wherein: Acquisition module: used to acquire the function of turbine blade structure, output observation dataset, probability density function of random input under given uncertainty distribution parameters, and prior uniform distribution probability density function of uncertainty distribution parameters; Construction module: used to construct the joint density function of the random input and uncertainty distribution parameters of the turbine blade structure based on the probability density function and the prior uniform distribution probability density function; First calculation module: used to introduce Dirac based on the joint density function and the function function. The function derivation derives the integral formula for the conditional density function of the output response at each output observation value. A joint distribution sample set of the joint density function is extracted using combined sampling techniques, and the corresponding function value is calculated. The Dirac delta method is employed. The normal approximation technique for the function, based on the joint distribution sample set and the corresponding function values, calculates the estimated value of the output response conditional density function at all output observation data values ​​through a single reliability analysis; The second calculation module is used to calculate the likelihood function of the uncertainty distribution parameter based on the estimated value of the conditional density function of the output response at all output observation data values ​​and the output observation dataset. The update module is used to construct a failure probability update formula for the turbine blade structure based on the function, the joint density function, and the likelihood function, and to calculate the failure probability of the updated turbine blade structure according to the failure probability update formula.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the turbine blade failure probability update method based on output observation data as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the turbine blade failure probability update method based on output observation data as described in any one of claims 1-7.