Multi-level calibration method for uncertainty parameters of finite element model of aircraft structure

Abstract

The invention relates to a multi-level calibration method for uncertainty parameters of an aircraft structure finite element model, which comprises the following steps of: 1, splitting the aircraft structure finite element model into a plurality of typical substructure finite element models, dividing the test data of the measuring points into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model; 2, screening uncertain parameters which are sensitive to the measurement point output influence of each typical substructure finite element model; step 3, calibrating the uncertainty parameter of each typical substructure finite element model by using corresponding calibration data by taking the uncertainty parameter which is sensitive to the measurement point output influence of each typical substructure finite element model as a variable; and 4, substituting the uncertainty parameters of each typical substructure finite element model into the aircraft structure finite element model, verifying by using verification data of the aircraft structure finite element model, and if the verification accuracy does not meet the requirement, carrying out the steps 2-4 again.

Description

Technical Field

[0001] The present application belongs to the technical field of aircraft structure finite element model parameter calibration, and specifically relates to a multi-level calibration method for uncertainty parameters of an aircraft structure finite element model. Background Art

[0002] Finite element model analysis is an important way to verify the strength of aircraft structures and an important method to obtain structural reliability.

[0003] The actual complex structure of an aircraft involves a large number of uncertain factors, such as material properties, geometric parameters, etc. Therefore, before using the finite element model to conduct a detailed analysis of the aircraft structure strength, it is often necessary to combine the test data of several working condition displacement / strain measurement points to calibrate the key uncertainty parameters of the finite element model to ensure the credibility of the finite element model.

[0004] However, aircraft structures that require strength verification often have complex structures and many uncertain parameters. The amount of test data for displacement / strain measurement points used for calibration is huge. It is inefficient and difficult to calibrate the uncertainty parameters of the entire finite element model using the test data of all measurement points at the same time.

[0005] This application is proposed in view of the above-mentioned technical defects. Summary of the invention

[0006] The purpose of the present application is to provide a multi-level calibration method for uncertainty parameters of an aircraft structure finite element model, so as to overcome or alleviate at least one of the known technical deficiencies.

[0007] The technical solution of this application is: A multi-level calibration method for uncertainty parameters of an aircraft structure finite element model, comprising: Step 1: split the aircraft structure finite element model into multiple typical substructure finite element models, and divide the test data of the measuring points into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model; Step 2: Screen the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point; Step 3: using the uncertainty parameters sensitive to the output of the measuring points of the finite element model of each typical substructure as variables and using the corresponding calibration data, calibrate the uncertainty parameters of the finite element model of each typical substructure; Step 4: Bring the uncertainty parameters of each typical substructure finite element model into the aircraft structure finite element model, and use the verification data of the aircraft structure finite element model for verification. If the verification accuracy does not meet the requirements, repeat steps 2 to 4.

[0008] Optionally, in the above-mentioned multi-level calibration method for uncertainty parameters of the aircraft structure finite element model, in step 2, each typical substructure finite element model measurement point outputs an uncertainty parameter that is sensitive to influence and is screened from the elastic modulus, Poisson's ratio and thickness of the structure.

[0009] Optionally, in the above-mentioned multi-level calibration method for uncertainty parameters of the aircraft structure finite element model, in step 2, a global Sobol index is used to perform sensitivity analysis on the uncertainty parameters to screen uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point.

[0010] Optionally, in the above-mentioned multi-level calibration method for uncertainty parameters of the aircraft structure finite element model, in step four, the output of each typical substructure finite element model measurement point and the calibration data are converted into Mahalanobis distance, and then the area measurement is calculated to quantitatively evaluate the consistency between the output of each typical substructure finite element model measurement point and the calibration data.

[0011] Optionally, in the above-mentioned multi-level calibration method for uncertainty parameters of the aircraft structure finite element model, in step 4, the uncertainty parameters of each typical substructure finite element model are calibrated based on the Bayesian optimization method.

[0012] This application has at least the following beneficial technical effects: A multi-level calibration method for uncertainty parameters of an aircraft structure finite element model is provided. The complex aircraft structure finite element model is split into multiple typical substructure finite element models, and the test data of the measuring points are divided into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model. Then, on the basis of screening the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measuring point, the Mahalanobis distance is used to reduce the dimension of the measuring point scale, the consistency of the multi-measurement point test data and the model output results is calculated based on a single indicator area metric, and the uncertainty parameters of each typical substructure finite element model are calibrated based on the Bayesian optimization method, and substituted into the aircraft structure finite element model for accuracy verification. The key parameters of the model can be calibrated based on the large-scale test data of the complex structure, and a finite element analysis model close to the actual aircraft structure response can be efficiently obtained, which effectively supports the aircraft structure strength verification. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a schematic diagram of a multi-level calibration method for uncertainty parameters of an aircraft structure finite element model provided in an embodiment of the present application; Figure 2 It is a schematic diagram of splitting the finite element model of the aircraft wing box section into the finite element models of the upper skin, the lower skin, the beam, and the rib provided in an embodiment of the present application; Figure 3It is a schematic diagram of the cumulative distribution function of the Mahalanobis distance between the measured point output of the finite element model of the aircraft wing box section and the verification data provided in the embodiment of the present application.

[0014] In order to better illustrate the present embodiment, some contents of the drawings may be omitted, enlarged or reduced, which is only used for illustrative purposes and should not be construed as limiting the present application. DETAILED DESCRIPTION

[0015] In order to make the technical solution and advantages of the present application clearer, the technical solution of the present application will be described in further detail in detail and in detail with reference to the accompanying drawings. It can be understood that the specific embodiments described here are only partial embodiments of the present application, which are only used to explain the present application, not to limit the present application. It should be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, and other related parts can refer to the general design.

[0016] In addition, unless otherwise defined, the technical terms or scientific terms used in the description of this application should be the common meanings understood by those skilled in the art in the field to which this application belongs. The term "include" used in the description of this application means that the concepts appearing before the term include the concepts listed after the term and their equivalents, without excluding other related concepts.

[0017] A multi-level calibration method for uncertainty parameters of aircraft structure finite element models, such as Figure 1 As shown in the figure, the complex aircraft structure finite element model is split into multiple typical substructure finite element models, and the scale of uncertainty multivariate consistency evaluation indicators is reduced. The uncertainty parameters of typical substructure finite element models and their aircraft structure finite element models are calibrated hierarchically, which can efficiently obtain a finite element analysis model that is close to the actual aircraft structure response, effectively supporting the aircraft structure strength verification.

[0018] Step 1: split the aircraft structure finite element model into multiple typical substructure finite element models, and divide the test data of the measuring points into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model.

[0019] Step 2: Screen the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point.

[0020] Write a program to modify the uncertainty parameters of each typical substructure finite element model and obtain a series of simulation results.

[0021] The global Sobol index is used to perform sensitivity analysis on uncertainty parameters, and the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point are screened, referring to the following formula: ; ; ; in: , For the , uncertainty parameter Output of measurement points for typical substructure finite element model Contribution to variance; For the , uncertainty parameter Interaction, output of measurement points for typical substructure finite element model Contribution to variance; For the uncertainty parameter Interacting with other uncertainty parameters, outputting the measured points of the finite element model of a typical substructure Contribution to variance; Output of measurement points for a typical substructure finite element model The total variance of For the uncertainty parameter , output of the measurement points of the typical substructure finite element model expectations; For the , uncertainty parameter , output of the measurement points of the typical substructure finite element model expectations; , for , The variance of is the first-order partial variance; is the second-order partial variance, By analogy Order partial variance.

[0022] Step 3: Using the uncertainty parameters that are sensitive to the output of the measuring points of each typical substructure finite element model as variables and using the corresponding calibration data, calibrate the uncertainty parameters of each typical substructure finite element model.

[0023] Output the measurement points of each typical substructure finite element model With calibration data , converted to Mahalanobis distance , and then calculate the area measure , quantitatively evaluate the output of each typical substructure finite element model measurement point With calibration data In order to ensure consistency, the uncertainty parameters of the finite element models of various typical substructures are calibrated based on the Bayesian optimization method.

[0024] ; ; ; in: , Output of measurement points for a typical substructure finite element model , calibration data expectations, represents the sample covariance matrix; is the Mahalanobis distance The cumulative distribution function of .

[0025] Step 4: Bring the uncertainty parameters of each typical substructure finite element model into the aircraft structure finite element model, and use the verification data of the aircraft structure finite element model for verification. If the verification accuracy does not meet the requirements, repeat steps 2 to 4.

[0026] For the evaluation of the verification accuracy of the aircraft structure finite element model, the consistency evaluation method for the calibration of uncertainty parameters of each typical substructure finite element model can be referred to, which will not be elaborated here.

[0027] In the multi-level calibration method for uncertainty parameters of an aircraft structure finite element model disclosed in the above embodiment, a complex aircraft structure finite element model is split into multiple typical substructure finite element models, and the test data of the measuring points are divided into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model. Then, on the basis of screening the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measuring point, the Mahalanobis distance is used to reduce the dimension of the measuring point scale, and the consistency of the multi-measurement point test data and the model output results is calculated based on a single indicator area metric. Based on the Bayesian optimization method, the uncertainty parameters of each typical substructure finite element model are calibrated and substituted into the aircraft structure finite element model for accuracy verification.

[0028] The multi-level calibration method for uncertainty parameters of the aircraft structure finite element model disclosed in the above embodiment can calibrate the key parameters of the model based on large-scale test data of complex structures, and can efficiently obtain a finite element analysis model that is close to the actual aircraft structure response, effectively supporting the aircraft structure strength verification.

[0029] In a specific example, the uncertainty parameters of the finite element model of an aircraft structure disclosed in the above embodiment are calibrated using the multi-level calibration method for uncertainty parameters of the finite element model of an aircraft structure, and the implementation is as follows.

[0030] S1. Split the finite element model of the aircraft wing box into multiple typical substructure finite element models, including upper skin, lower skin, beam, and rib finite element models.

[0031] The test data of the measuring points are the test measurement data of 96 strain gauges, which are divided into five groups of data: A, B, C, D, and E. The specific groups are as follows:

[0032] S2. Uncertainty parameters of each typical substructure finite element model are selected: elastic modulus E, Poisson's ratio , thickness t, and use the global Sobol index to perform sensitivity analysis on the uncertainty parameters, and screen the uncertainty parameters that are sensitive to the output of the finite element model measurement points of each typical substructure. Among them, the uncertainty parameters that are sensitive to the output of the finite element model measurement points of the upper skin are the elastic modulus E1 and Poisson's ratio of the upper skin material. , thickness t1; the uncertainty parameters that are sensitive to the output of the lower skin finite element model measurement point are the elastic modulus E2 of the lower skin material, Poisson's ratio , thickness t2; the uncertainty parameter that is sensitive to the output of the beam finite element model measuring point is the thickness t3 of the beam structure; the uncertainty parameter that is sensitive to the output of the rib finite element model measuring point is the thickness t4 of the rib structure.

[0033] S3. Taking the uncertainty parameters that are sensitive to the output of the upper skin, lower skin, beam, and rib finite element model measurement points as variables, the uncertainty parameters of the upper skin, lower skin, beam, and rib finite element models are calibrated using four sets of calibration data A, B, C, and D to obtain the optimal parameter distribution, as shown in the following table:

[0034] S4, the uncertainty parameters E1, ,t1,E2, , t2, t3, t4, are brought into the finite element model of the aircraft wing box section and verified using the E group verification data. The cumulative distribution function of the Mahalanobis distance between the measurement point output and the verification data is as follows: Figure 3 As shown, the error is small and the accuracy meets the requirements. The finite element model of the aircraft wing box section is constructed in this way, which can effectively support the strength verification of the aircraft wing box section.

[0035] So far, the technical solution of the present application has been described in conjunction with the preferred embodiments shown in the accompanying drawings. Those skilled in the art should understand that the scope of protection of the present application is obviously not limited to these specific embodiments. Without departing from the principles of the present application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the scope of protection of the present application.

Claims

1. A multi-level calibration method for uncertainty parameters of an aircraft structure finite element model, characterized in that: include: Step 1: splitting the aircraft structure finite element model into multiple typical substructure finite element models, and dividing the test data of the measuring points into calibration data of each typical substructure finite element model and verification data of the aircraft structure finite element model; Step 2: Screen the uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point; Step 3: using the uncertainty parameters sensitive to the output of the measuring points of the finite element model of each typical substructure as variables and using the corresponding calibration data, calibrate the uncertainty parameters of the finite element model of each typical substructure; Step 4: Bring the uncertainty parameters of each typical substructure finite element model into the aircraft structure finite element model, and use the verification data of the aircraft structure finite element model for verification. If the verification accuracy does not meet the requirements, repeat steps 2 to 4.

2. The multi-level calibration method for uncertainty parameters of an aircraft structure finite element model according to claim 1, characterized in that: In step 2, the output of each typical substructure finite element model measurement point affects sensitive uncertainty parameters, which are selected from the elastic modulus, Poisson's ratio, and thickness of the structure.

3. The multi-level calibration method for uncertainty parameters of an aircraft structure finite element model according to claim 2, characterized in that: In step 2, the global Sobol index is used to perform sensitivity analysis on uncertainty parameters to screen out uncertainty parameters that are sensitive to the output of each typical substructure finite element model measurement point.

4. The method for multi-level calibration of uncertainty parameters of an aircraft structure finite element model according to claim 3, characterized in that: In step 4, the output of each typical substructure finite element model measurement point and the calibration data are converted into Mahalanobis distance, and then the area measurement is calculated to quantitatively evaluate the consistency between the output of each typical substructure finite element model measurement point and the calibration data.

5. The method for multi-level calibration of uncertainty parameters of an aircraft structure finite element model according to claim 4, characterized in that: In step 4, the uncertainty parameters of each typical substructure finite element model are calibrated based on the Bayesian optimization method.

Images