Method for correcting composite material model in dynamic environment

Abstract

According to the method for correcting the composite material model in the dynamic environment, the transmission relation between the physical property parameters and the dynamic modal parameters of the composite material is established, and the physical property parameters of a structure are corrected, so that the error between a finite element model and a test model is reduced, and the precision of the finite element model is remarkably improved; the proxy model constructed by using the BP neural network does not need to depend on a mass matrix and a stiffness matrix of the system, and parameter reverse identification can be realized completely based on a data-driven transfer relationship. According to the method, the time required for model correction is greatly shortened, and the correction efficiency is remarkably improved.

Description

Technical Field

[0001] The invention belongs to the technical field of composite material model correction, and in particular relates to a composite material model correction method under a dynamic environment. Background Art

[0002] Composite material model modification under dynamic environment is an important scientific issue brought out by aerospace engineering. It is a key technology guiding lightweight design of structures and ground assessment tests. It is an important basis for conducting coupled field response analysis of structures and predicting service performance and safety performance.

[0003] At present, the finite element analysis method of typical composite materials has been widely used in the aerospace field. However, it is almost impossible to establish a finite element model that can accurately describe the actual object based solely on engineering experience. There are various sources of errors in finite element modeling. Mottershead and other scholars have summarized the model errors into three main aspects and discussed them in detail: structural error, parameter error and order error. Among them, the structural error refers to the rationality of the selected mathematical model, and the order error is related to the degree of discretization. These two types of errors can usually be avoided through reasonable modeling. In contrast, parameter errors caused by differences in material parameters, boundary conditions, etc. often become the focus of finite element model correction.

[0004] The design parameter model correction method constructs an objective function and repeatedly iterates the calculation of the correction amount, thereby effectively reducing the error. This method is applicable to the physical parameters, geometric parameters and boundary conditions of the correction model, and is not limited by the correction object. The corrected parameters have clear physical meanings and can be directly compared with the actual situation, so they have been widely used in engineering practice. However, the sensitivity-based finite element model correction method requires repeated iterative calculations of the finite element model, which is often a time-consuming process.

[0005] In view of the problem that the existing design parameter model correction method takes a long time in the repeated iteration process of the finite element model, it is urgent to explore effective ways to shorten the correction time in order to improve the efficiency of the correction process. Summary of the invention

[0006] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a composite material model correction method under a dynamic environment. The present invention can solve the problems existing in the prior art.

[0007] The technical solution of the present invention:

[0008] A composite material model correction method under a dynamic environment comprises the following steps:

[0009] Step 1: Conduct structural modal tests on typical composite materials and random vibration tests at room temperature or high temperature to obtain dynamic modal parameters of typical structures;

[0010] Step 2: Establish a fine finite element model, estimate the range of physical property parameters of composite materials based on engineering experience, and use Latin hypercube sampling to input the sample parameters into the fine finite element model of the typical structure of the composite material to obtain the dynamic modal parameters of the skin-reinforced structure under different physical property parameters, and establish the sample space of physical property parameters and corresponding modal parameters;

[0011] Step 3: train the BP neural network proxy model through the sample space, take the physical property parameters as input and the modal parameters as output to establish a mapping relationship between the two. The BP neural network proxy model fits the data in two directions: forward propagation and reverse propagation. Forward propagation refers to the information being transmitted from the input layer to the output layer in sequence to generate output results; reverse propagation refers to the neural network adjusting and modifying the weights and thresholds layer by layer by reversely transmitting errors.

[0012] Step 4: Based on the BP neural network proxy model obtained after training, the minimum norm of the modal parameters output by different physical parameters and the modal parameters obtained by the test is used as the optimization target, and the optimization algorithm is used to optimize and obtain the optimal solution of the modal parameters of the typical structure of the composite material, thereby completing the correction of the fine finite element model and obtaining the typical structural mechanics model of the composite material;

[0013] Step five, by establishing the BP neural network proxy model under different working conditions, the physical properties of the material corresponding to different temperatures are obtained, and the physical properties of the material at different temperature moments are fitted using the least squares method, so that the material parameters under the temperature conditions where no test is conducted can be predicted, and a typical structural mechanics model of the composite material under a dynamic environment can be obtained.

[0014] Furthermore, the method for obtaining the optimal solution of the dynamic modal parameters of a typical structure of a composite material is:

[0015] The goal of model correction is to minimize the residual value between the real dynamic modal information parameters obtained from the structural modal test and the dynamic modal parameters calculated by the proxy model. The model correction problem is transformed into an optimization problem of minimizing the dynamic modal frequency residual, and then the minimum value problem is solved by the optimization method.

[0016] Furthermore, the BP neural network agent model includes an input layer, a hidden layer and an output layer. The forward propagation of information is from the input layer to the hidden layer and then to the output layer, and the reverse propagation of the error is from the output layer to the hidden layer and then to the input layer. The forward propagation of information and the reverse propagation of error are repeated alternately and cyclically until the error meets the pre-set requirements, or the number of neural network training times reaches the maximum value, the optimization process ends, and the optimal solution is output.

[0017] Furthermore, in step three, the input physical property parameters are normalized, that is, the data in each column is divided by the maximum value of all the values ​​in the column, and the normalized value is used as the input physical property parameter.

[0018] Furthermore, in step 4, the optimal solution is obtained and then the inverse process of the normalization process is performed, and the optimal solution is restored to the original numerical value to obtain the optimized material property parameters.

[0019] Furthermore, the BP neural network proxy model uses the nonlinear encoding characteristics of neural networks to establish a nonlinear mapping from physical parameters to modal parameters.

[0020] The beneficial effects of the present invention compared with the prior art are as follows:

[0021] 1) The present invention reasonably normalizes the data so that the established proxy model does not impose any restrictions on the numerical values ​​of parameters such as the elastic modulus, shear modulus, Poisson's ratio and thermal expansion coefficient in the finite element model, which significantly improves the applicability and flexibility of the correction method and realizes the unconstrained selection of correction parameters.

[0022] 2) The present invention utilizes the characteristics of BP neural network to perform nonlinear distributed coding on input and output parameters, and constructs a nonlinear mapping relationship from input to output. When the network scale is large enough, information is embedded in the network weights to form a highly redundant information storage mode, thereby ensuring that all influencing factors can be encoded simultaneously. Due to its integrity, dynamics and correlation, BP neural network has strong fault tolerance, thereby significantly improving the accuracy of model correction.

[0023] 3) The proxy model constructed by the BP neural network in the present invention does not need to rely on the mass matrix and stiffness matrix of the system, and can achieve parameter reverse identification based entirely on the data-driven transfer relationship. This method greatly shortens the time required for model correction and significantly improves the correction efficiency.

[0024] 4) The present invention establishes a transfer relationship between the physical property parameters of the composite material and the dynamic modal parameters to correct the physical property parameters of the structure, thereby reducing the error between the finite element model and the test model and significantly improving the accuracy of the finite element model. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The included drawings are used to provide a further understanding of the embodiments of the present invention, which constitute a part of the specification, are used to illustrate the embodiments of the present invention, and together with the text description, explain the principles of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 It shows a technical route of a composite material model correction method under a dynamic environment provided by an embodiment of the present invention;

[0027] Figure 2 A schematic diagram of a typical skin reinforcement structure provided according to an embodiment of the present invention is shown;

[0028] Figure 3 A network schematic diagram of a BP neural network agent model provided according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0029] It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0031] Unless otherwise specifically stated, the relative arrangement, numerical expressions and numerical values ​​of the parts and steps set forth in these embodiments do not limit the scope of the present invention. Meanwhile, it should be understood that, for ease of description, the sizes of the various parts shown in the accompanying drawings are not drawn according to the actual proportional relationship. The technology, method and equipment known to those of ordinary skill in the relevant art may not be discussed in detail, but in appropriate cases, the technology, method and equipment should be considered as a part of the specification. In all examples shown and discussed here, any specific value should be interpreted as being merely exemplary, rather than as a limitation. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, and therefore, once a certain item is defined in an accompanying drawing, it does not need to be further discussed in subsequent drawings.

[0032] like Figure 1 As shown, a composite material model correction method in a dynamic environment includes the following steps:

[0033] Step 1: Conduct structural modal tests on typical composite materials and random vibration tests at room temperature or high temperature to obtain dynamic modal parameters of typical composite materials.

[0034] Step 2: Establish a fine finite element model, estimate the range of physical property parameters of composite materials based on engineering experience, and input the sample parameters into the fine finite element model of the typical structure of the composite material through Latin hypercube sampling to obtain the dynamic modal parameters of the typical structure of the composite material under different physical property parameters, and establish the sample space of physical property parameters and corresponding modal parameters;

[0035] Step 3: train the BP neural network proxy model through the sample space, take the physical property parameters as input and the modal parameters as output to establish a mapping relationship between the two. The BP neural network proxy model fits the data in two directions: forward propagation and reverse propagation. Forward propagation refers to the information being transmitted from the input layer to the output layer in sequence to generate output results; reverse propagation refers to the neural network adjusting and modifying the weights and thresholds layer by layer by reversely transmitting errors.

[0036] Step 4: Based on the BP neural network proxy model obtained after training, the minimum norm of the modal parameters output by different physical parameters and the modal parameters obtained by the experiment is used as the optimization target. The optimization is performed through the optimization algorithm to obtain the optimal solution of the typical structural modal parameters of the composite material, thereby completing the correction of the fine finite element model and obtaining the typical structural mechanics model of the composite material under a dynamic environment.

[0037] Furthermore, the method for obtaining the optimal solution of the dynamic modal parameters of a typical structure of a composite material is:

[0038] The goal of model correction is to minimize the residual value between the real dynamic modal information parameters obtained from the structural modal test and the dynamic modal parameters calculated by the proxy model. The model correction problem is transformed into an optimization problem of minimizing the dynamic modal frequency residual, and then the minimum value problem is solved by the optimization method.

[0039] Furthermore, in one embodiment, the BP neural network agent model includes an input layer, a hidden layer and an output layer. The forward propagation of information is from the input layer to the hidden layer and then to the output layer, and the reverse propagation of the error is from the output layer to the hidden layer and then to the input layer. The forward propagation of information and the reverse propagation of error are repeated alternately and cyclically until the error meets the pre-set requirements, or the number of neural network training times reaches the maximum value, the optimization process ends, and the optimal solution is output.

[0040] Furthermore, in one embodiment, in step 3, the input physical property parameters are normalized, that is, the data in each column is divided by the maximum value of all the values ​​in the column, and the normalized value is used as the input physical property parameter. By reasonably normalizing the data, the established proxy model does not set any restrictions on the numerical values ​​of parameters such as elastic modulus, shear modulus, Poisson's ratio and thermal expansion coefficient in the finite element model, which significantly improves the applicability and flexibility of the correction method and realizes the unconstrained selection of correction parameters.

[0041] Furthermore, in one embodiment, in step 4, the optimal solution is obtained and then the inverse process of the normalization process is performed to restore the optimal solution to the original numerical value to obtain the optimized material property parameters.

[0042] Further in one embodiment, the BP neural network proxy model uses the characteristics of nonlinear coding of neural networks to establish a nonlinear mapping of physical parameters to modal parameters. The characteristics of nonlinear distributed coding of input and output parameters by the BP neural network are used to construct a nonlinear mapping relationship from input to output. When the network scale is large enough, information is embedded in the network weights to form a highly redundant information storage mode, thereby ensuring that all influencing factors are encoded at the same time. Due to its integrity, dynamics and correlation, the BP neural network has a strong fault tolerance ability, thereby significantly improving the accuracy of model correction.

[0043] In order to have a further understanding of the typical structural mechanics modeling method of composite materials under dynamic conditions provided by the present invention, a detailed description is given below in conjunction with specific examples and drawings.

[0044] Take the typical composite skin reinforcement structure as an example:

[0045] In this embodiment, the typical skin reinforcement structure is as follows: Figure 2As shown, the typical skin reinforcement structure includes two parts: skin and reinforcement, which are made of C / SiC composite material. The material properties at room temperature are elastic modulus E1=E2=55GPa, shear modulus G 12 =100GPa, G 13 =G 23 =20GPa, isotropic Poisson's ratio μ=0.05, isotropic thermal expansion coefficient α=1.3E-6.

[0046] The composite material model correction method in a dynamic environment first determines E1, E2, G according to engineering experience. 12 , α is the material property parameter to be optimized, and the value range of the parameter to be optimized is estimated. In this embodiment, the elastic modulus E1=E2 has a value range of 40-80GPa, and the shear modulus G 12 The value range is 80-150GPa, and the value range of thermal expansion coefficient α is 1E-6-2E-6.

[0047] The composite material model correction method under dynamic environment, after determining the value range and the parameters to be optimized, establishes a sample space within the parameter value range through Latin hypercube sampling. In this embodiment, a sample space containing 600 sets of data is established.

[0048] The composite material model correction method under a dynamic environment, after completing sampling, establishes a finite element model through existing technology, calculates and obtains the modal parameters of each sample point in the sample space, and establishes a training set for training the BP neural network proxy model. In one embodiment, in order to improve the calculation efficiency, the skin-reinforced structure establishes a finite element model through shell units.

[0049] In the composite material model correction method under dynamic environment, the selection of the number of hidden layer nodes is the most uncertain part of the entire proxy model. In this embodiment, the number of hidden layer nodes is determined to be 10 by empirical formula. The constructed neural network is as follows: Figure 3 shown.

[0050] In order to achieve better training and fitting effects, the composite material model correction method under a dynamic environment performs normalization on the training data before training, and divides the data in each column by the maximum value of all values ​​in the column, so as to solve the problem of different sensitivities of different parameters in the fitting process caused by large differences in the magnitudes of various parameters.

[0051] The composite material model correction method under a dynamic environment selects 400 groups of data as training sets and 200 groups of data as test sets after normalizing the data, uses physical property parameters as input modes and parameters as output to train and verify the BP neural network proxy model, and uses the characteristics of nonlinear coding of the neural network to establish a nonlinear mapping from physical property parameters to modal parameters to improve the prediction accuracy of the proxy model.

[0052] The composite material model correction method under a dynamic environment, after completing the construction of the proxy model, takes the minimum residual value between the real dynamic modal information parameters obtained by the structural modal test and the dynamic modal parameters calculated by the proxy model as the goal of model correction, converts the model correction problem into an optimization problem of minimizing the dynamic modal frequency residual, and then solves the minimum value problem through the particle swarm optimization algorithm. After obtaining the optimal solution, the inverse process of the normalization process is performed, and the optimal solution is restored to the original value to obtain the optimized material property parameters.

[0053] The composite material model correction method under a dynamic environment obtains the dynamic parameters of the structure under different temperature conditions by conducting dynamic experiments under different temperature conditions, and obtains the material property parameters corresponding to different temperatures by establishing the BP neural network proxy model under different conditions. The material property parameters obtained at different temperature moments are fitted using the least squares method, so that the material parameters under the temperature conditions where no test is conducted can be predicted.

[0054] In summary, the present invention provides at least the following advantages compared to the prior art:

[0055] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A composite material model correction method under dynamic environment, characterized in that: The following steps are involved: Step 1: Conduct structural modal tests on the composite skin-reinforced structure, perform random vibration tests under normal or high temperature conditions, and obtain dynamic modal parameters of the skin-reinforced structure; Step 2: Establish a fine finite element model, estimate the range of physical property parameters of composite materials based on engineering experience, and input the sample parameters into the fine finite element model of the composite skin-reinforced structure through Latin hypercube sampling to obtain the dynamic modal parameters of the skin-reinforced structure under different physical property parameters, and establish the sample space of physical property parameters and corresponding modal parameters; Step 3: train the BP neural network proxy model through the sample space, take the physical property parameters as input and the modal parameters as output to establish a mapping relationship between the two. The BP neural network proxy model fits the data in two directions: forward propagation and reverse propagation. Forward propagation refers to the information being transmitted from the input layer to the output layer in sequence to generate output results; reverse propagation refers to the neural network adjusting and modifying the weights and thresholds layer by layer by reversely transmitting errors. Step 4: Based on the BP neural network proxy model obtained after training, the minimum norm of the modal parameters output by different physical parameters and the modal parameters obtained by the test is used as the optimization target, and the optimization is performed through the optimization algorithm to obtain the optimal solution of the modal parameters of the skin-reinforced structure, thereby completing the correction of the fine finite element model and obtaining the mechanical model of the skin-reinforced structure; Step five, by establishing the BP neural network proxy model under different working conditions, the physical properties of the material corresponding to different temperatures are obtained, and the physical properties of the material at different temperature moments are fitted using the least squares method. The material parameters under the temperature conditions not tested can be predicted, and the mechanical model of the skin reinforcement structure under a dynamic environment can be obtained.

2. The composite material model correction method under dynamic environment according to claim 1, characterized in that: The method for obtaining the optimal solution of the dynamic modal parameters of the skin-reinforced structure is: The goal of model correction is to minimize the residual value between the real dynamic modal information parameters obtained from the structural modal test and the dynamic modal parameters calculated by the proxy model. The model correction problem is transformed into an optimization problem of minimizing the dynamic modal frequency residual, and then the minimum value problem is solved by the optimization method.

3. The composite material model correction method under dynamic environment according to claim 2, characterized in that: The BP neural network agent model includes an input layer, a hidden layer and an output layer. The forward propagation of information is from the input layer to the hidden layer and then to the output layer. The reverse propagation of the error is from the output layer to the hidden layer and then to the input layer. The forward propagation of information and the reverse propagation of the error are repeated alternately and cyclically until the error meets the pre-set requirements or the number of neural network training times reaches the maximum value, the optimization process ends, and the optimal solution is output.

4. The composite material model correction method under dynamic environment according to claim 1, characterized in that: In step three, the input physical property parameters are normalized, that is, the data in each column is divided by the maximum value of all the values ​​in the column, and the normalized value is used as the input physical property parameter.

5. The composite material model correction method under dynamic environment according to claim 4, characterized in that: In step 4, the optimal solution is obtained and then the inverse process of the normalization process is performed to restore the optimal solution to the original numerical value to obtain the optimized material property parameters.

6. The composite material model correction method under dynamic environment according to claim 3, characterized in that: The BP neural network proxy model uses the characteristics of nonlinear coding of neural networks to establish a nonlinear mapping from physical parameters to modal parameters.

Images