FSW joint fatigue crack propagation life prediction method based on PSO-SVR

Abstract

The invention discloses a PSO-SVR-based FSW joint fatigue crack propagation life prediction method and electronic equipment, and the method comprises the steps: simulating a crack propagation process of an FSW joint through a finite element simulation technology, obtaining a crack tip stress intensity factor, a corresponding loading load size and a crack length of the FSW joint in the process as a group of data, and constructing a data set; performing standardization processing on the constructed data set, and then dividing the data set into a training set and a test set; and training a machine learning model by using the training set, evaluating the trained model by using the test set, and performing hyper-parameter optimization on the model until an optimal crack tip stress intensity factor prediction model is obtained. According to the PSO-SVR-based FSW joint fatigue crack propagation life prediction method and the electronic equipment provided by the embodiment of the invention, finite element simulation data and test data are fused, so that the time cost and expense for constructing the fatigue crack propagation life model are effectively reduced, and the method and the electronic equipment have the advantages of high calculation efficiency, high prediction precision and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of prediction of material fatigue life, and particularly relates to a method for predicting the fatigue crack growth life of a FSW joint based on PSO-SVR and an electronic device. Background Art

[0002] Compared with traditional TIG and MIG fusion welding, the aluminum-lithium alloy welded by friction stir welding (FSW) has higher performance and does not generate defects such as pores and inclusions. During the crack propagation process, the FSW joint will form a large stress concentration under stress, which will further affect the crack growth life of the material. Therefore, predicting the fatigue crack growth life of welded structures has important engineering application value for the reliability assessment and damage tolerance design of aerospace equipment.

[0003] In related technologies, machine learning algorithms driven by data are generally used to study the fatigue crack growth life of welded structures. However, due to problems such as the high difficulty and cost of fatigue crack growth tests for FSW welded joints, the amount of test data is small, and it is difficult to use the test data as a dataset for machine learning to train a model that can accurately predict the life of FSW welded structures. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems in the related technologies to some extent. For this reason, the purpose of the present invention is to propose a method for predicting the fatigue crack growth life of a FSW joint based on PSO-SVR and an electronic device to improve the life prediction efficiency of FSW welded structures and reduce costs.

[0005] To achieve the above object, a first aspect embodiment of the present invention proposes a method for predicting the fatigue crack growth life of a FSW joint based on PSO-SVR, and the method for predicting the fatigue crack growth life of the FSW joint includes:

[0006] Simulate the crack propagation process of the FSW joint through finite element simulation technology, and obtain the stress intensity factor at the crack tip, the corresponding loading load magnitude, and the crack length of the FSW joint during the process as a set of data to construct a dataset;

[0007] Perform standardization processing on the constructed dataset, and then divide the dataset into a training set and a test set;

[0008] Train a machine learning model with the training set, evaluate the trained model with the test set, and optimize the hyperparameters of the model at the same time until an optimal stress intensity factor prediction model at the crack tip is obtained;

[0009] Initialize the relevant parameters of crack propagation;

[0010] Based on the fatigue crack growth theory in linear elastic fracture mechanics, the fatigue crack growth is carried out by combining the stress intensity factor prediction model at the crack tip and the crack growth rate model, and the predicted value of the fatigue crack growth life of the FSW joint is obtained.

[0011] According to an embodiment of the present invention, the process of simulating the crack growth process of the FSW joint by finite element simulation technology includes:

[0012] Partitioned modeling is used to simulate the material properties of the FSW welding part and the base metal part.

[0013] According to an embodiment of the present invention, the steps of normalizing the constructed data set include:

[0014]

[0015] In the formula, is the mean of the feature data; σ(x) is the standard deviation of the feature data; X i is the data after normalization.

[0016] According to an embodiment of the present invention, the normalized data set is divided into a training set and a test set, where the training set accounts for 75% and the test set accounts for 25%.

[0017] According to an embodiment of the present invention, the machine learning model is a support vector regression model, and its kernel function uses the RBF kernel function. Among them, the hyperparameters of the model include:

[0018] Penalty factor C, insensitive loss factor ε, and kernel parameter γ.

[0019] According to an embodiment of the present invention, the method for optimizing the hyperparameters of the machine learning model includes:

[0020] The particle swarm optimization algorithm is used to optimize the penalty factor C, the insensitive loss factor ε, and the kernel parameter γ.

[0021] According to an embodiment of the present invention, the trained model is evaluated using the test set, and the evaluation methods include:

[0022] The coefficient of determination R 2 and the mean squared error MSE are used as the criteria for evaluating the model, where:

[0023]

[0024] Among them, N represents the total number of samples; y i represents the true value of the model; represents the predicted value of the model; represents the mean of the true values.

[0025] According to an embodiment of the present invention, the step of initializing the relevant parameters for crack propagation includes:

[0026] Set the initial parameters for crack propagation, including the initial crack length, crack propagation direction, and loading conditions, according to the actual working conditions of the FSW joint and the initial crack state.

[0027] According to an embodiment of the present invention, the fatigue crack propagation is jointly extended by a crack propagation rate model and a crack tip stress intensity factor prediction model:

[0028]

[0029] where a k is the crack length at the k-th crack propagation step; Δa is the crack propagation growth step; N k is the fatigue life at the k-th crack propagation step; is the crack propagation rate at a k-1 under.

[0030] To achieve the above object, an embodiment of the second aspect of the present invention proposes an electronic device, including a memory, a processor, and a computer program stored on the memory. When the computer program is executed by the processor, the above-mentioned PSO-SVR-based FSW joint fatigue crack propagation life prediction method is implemented.

[0031] The PSO-SVR-based FSW joint fatigue crack propagation life prediction method and electronic device according to the embodiments of the present invention effectively reduce the time cost and cost of constructing the fatigue crack propagation life model by fusing finite element simulation data and experimental data, and have the advantages of high calculation efficiency, low cost, high prediction accuracy, etc., and can be used for crack propagation life prediction of FSW joints under various working conditions such as different loads and initial crack lengths. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a construction flow chart of the PSO-SVR-based FSW joint fatigue crack propagation life prediction method according to an embodiment of the present invention;

[0033] Figure 2 is a precision comparison chart of the fatigue crack propagation life curve predicted by the PSO-SVR-based FSW joint fatigue crack propagation life prediction method according to an embodiment of the present invention;

[0034] Figure 3 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where like or similar reference numerals denote like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as limiting the present invention.

[0036] The following describes a method and an electronic device for predicting the fatigue crack growth life of FSW joints based on PSO-SVR according to embodiments of the present invention with reference to the accompanying drawings.

[0037] Figure 1 It is a construction flowchart of a method for predicting the fatigue crack growth life of FSW joints based on PSO-SVR according to an embodiment of the present invention.

[0038] As Figure 1 shown, the method for predicting the fatigue crack growth life of FSW joints based on PSO-SVR includes:

[0039] S1. Simulate the crack growth process of the FSW joint through finite element simulation technology, and obtain the stress intensity factor at the crack tip of the FSW joint, the corresponding loading load magnitude, and the crack length during the process as a set of data to construct a data set.

[0040] As an example, during the process of simulating the crack growth of the FSW joint through finite element simulation technology, the material properties of the FSW welding part and the base metal part are simulated by using partitioned modeling. This is to accurately simulate the mechanical behavior changes caused by the material property differences between the FSW welding part and the base metal part. This partitioned modeling strategy ensures the accuracy of the simulation results.

[0041] S2. Perform standardization processing on the constructed data set, and then divide the standardized data set into a training set and a test set.

[0042] Specifically, the training set accounts for 75%, and the test set accounts for 25%.

[0043] S3. Train a machine learning model with the above training set, and continuously adjust the internal parameters of the model so that it can accurately learn from the data and predict the stress intensity factor at the crack tip; evaluate the trained model with the above test set to test its prediction performance on unseen data. During the evaluation process, at the same time, implement a hyperparameter optimization strategy for the model, and iteratively adjust the hyperparameters of the model, such as the learning rate, regularization strength, kernel function parameters, etc., in order to find the best parameter combination to minimize the prediction error of the model on the test set. This process continues until a prediction model of the stress intensity factor at the crack tip with the best performance on the test set is obtained;

[0044] S4. Initialize the relevant parameters of the crack growth.

[0045] Specifically, the steps of initializing the relevant parameters of crack propagation include:

[0046] According to the actual working conditions of the FSW joint and the initial crack state, set the initial parameters of crack propagation, including the initial crack length, crack propagation direction, loading conditions, etc.

[0047] S5. Based on the classical theory of fatigue crack propagation in linear elastic fracture mechanics, combine the crack tip stress intensity factor prediction model obtained by the above training and optimization and the mathematical model suitable for describing the crack propagation rate to construct a comprehensive model that can simulate the crack propagation process of the FSW joint under given loading conditions. This model can dynamically predict the change of the crack tip stress intensity factor, calculate the crack propagation rate accordingly, and then predict the fatigue crack propagation life of the FSW joint through iterative calculation.

[0048] In some embodiments of the present invention, the steps of standardizing the constructed data set include:

[0049]

[0050] In the formula, is the mean value of the feature data; σ(x) is the standard deviation of the feature data; X i is the standardized data.

[0051] In some embodiments of the present invention, the machine learning model trained by the above training set is a support vector regression model, and its kernel function adopts the RBF kernel function. Among them, the hyperparameters of this model include: penalty factor C, insensitive loss factor ε, and kernel parameter γ.

[0052] Then, the hyperparameters of the above machine learning model can be optimized. The specific method includes: using the particle swarm optimization algorithm to optimize the penalty factor C, insensitive loss factor ε, and kernel parameter γ.

[0053] In some embodiments of the present invention, the trained model is evaluated with the test set. The evaluation method includes: using the coefficient of determination R 2 and the mean squared error MSE as the criteria for evaluating the model. Among them:

[0054]

[0055] Among them, N represents the total number of samples; y i represents the true value of the model; represents the predicted value of the model; represents the mean value of the true values. MSE reflects the difference between the predicted value and the true value of the model. The closer the value is to 0, the higher the accuracy of the model. R 2It reflects the correlation between the predicted value and the true value of the model. The closer the value is to 1, the higher the model accuracy.

[0056] Specifically, the dimension of the particle position is set to 3, and the size of the particle population is set to 20 as the initial conditions for the particle swarm optimization algorithm when optimizing the above hyperparameters. After 20 iterations, the optimal model can be obtained, with R 2 = 0.9995 and MSE = 36.5 on the test set.

[0057] As Figure 2 shown, the crack tip stress intensity factor prediction model and the crack growth rate model obtained above are fused to obtain a comprehensive model that can simulate the crack growth process of the FSW joint under given loading conditions. The comprehensive model can be expressed by the following formula:

[0058]

[0059] In the formula, a k is the crack length at the k-th crack growth step; Δa is the crack growth increment; N k is the fatigue life at the k-th crack growth step; is the crack growth rate at a k-1 The following.

[0060] Specifically, when crack growth occurs, first, the crack growth data is initialized and the initial crack is inserted; then, the crack tip stress intensity factor prediction model calculates the corresponding crack tip stress intensity factor according to the load magnitude and crack length; then, the crack growth rate model is used to calculate the crack growth rate, and crack growth is carried out through the following formula until the stress intensity factor reaches the fracture toughness; finally, the complete crack growth life of the FSW joint is obtained.

[0061] Corresponding to the above embodiments, as Figure 3 shown, a second aspect embodiment of the present invention proposes an electronic device. The electronic device 200 includes: a processor 201 and a memory 203. Among them, the processor 201 and the memory 203 are connected, such as through a bus 202. Optionally, the electronic device 200 may further include a transceiver 204. It should be noted that in practical applications, the transceiver 204 is not limited to one, and the structure of the electronic device 200 does not constitute a limitation to the embodiments of the present invention.

[0062] The processor 201 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of the present invention. The processor 201 can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0063] The bus 202 can include a path for transmitting information between the above components. The bus 202 can be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, etc. The bus 202 can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 2 only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0064] The memory 203 is used to store a computer program corresponding to the PSO-SVR-based FSW joint fatigue crack growth life prediction method of the foregoing embodiments of the present invention, and the execution of this computer program is controlled by the processor 201. The processor 201 is used to execute the computer program stored in the memory 203 to implement the content shown in the foregoing method embodiments.

[0065] Among them, the electronic device 200 includes but is not limited to: mobile terminals such as laptop computers, PADs (tablet computers), etc., and fixed terminals such as desktop computers, etc. Figure 2 The illustrated electronic device 200 is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present invention.

[0066] The electronic device 200 according to the embodiments of the present invention can effectively reduce the time cost and expenses for constructing a fatigue crack growth life model by fusing finite element simulation data and test data, and has the advantages of high computing efficiency, low cost, high prediction accuracy, etc., and can be used for predicting the crack growth life of FSW joints under various working conditions such as different loads and initial crack lengths.

[0067] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion having one or more wirings (electronic device), a portable computer diskette case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0068] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0069] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0070] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0071] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for predicting fatigue crack growth life of FSW joints based on PSO-SVR, characterized in that: The method comprises: The crack propagation process of the FSW joint is simulated by finite element simulation technology, and the crack tip stress intensity factor, the corresponding loading load size and the crack length of the FSW joint are obtained as a set of data to construct a data set; The constructed data set is standardized and then divided into a training set and a test set; The machine learning model is trained with the training set, and the trained model is evaluated with the test set. The hyperparameters of the model are optimized at the same time until the optimal crack tip stress intensity factor prediction model is obtained. Initialize the relevant parameters of crack extension; Based on the fatigue crack growth theory in linear elastic fracture mechanics, the crack tip stress intensity factor prediction model and the crack growth rate model are combined to carry out fatigue crack growth and obtain the fatigue crack growth life prediction value of FSW joints.

2. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1 is characterized in that: The method of simulating the crack propagation process of the FSW joint by using the finite element simulation technology includes: Partition modeling is used to simulate the material properties of the FSW welding area and the base material area.

3. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1 is characterized in that: The step of standardizing the constructed data set includes: In the formula, is the mean of the characteristic data; σ(x) is the standard deviation of the characteristic data; X i is the standardized data.

4. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1 is characterized in that ,The standardized data set is divided into a training set and a test set, where the training set accounts for 75% and the test set accounts for 25%.

5. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1, characterized in that: The machine learning model is a support vector regression model, and its kernel function adopts the RBF kernel function, wherein the hyperparameters of the model include: Penalty factor C, insensitive loss factor ε and kernel parameter γ.

6. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 5, characterized in that: The method for optimizing hyperparameters of a machine learning model comprises: The particle swarm optimization algorithm is used to optimize the penalty factor C, insensitive loss factor ε and kernel parameter γ.

7. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1, characterized in that: The training model is evaluated using the test set, and the evaluation method includes: The coefficient of determination R 2 And mean square error MSE as the criterion for evaluating the model, where: Where N is the total number of samples; y i Represents the true value of the model; represents the predicted value of the model; Represents the mean of the true values.

8. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1, characterized in that: The step of initializing parameters related to crack extension includes: According to the actual working conditions of the FSW joint and the initial state of the crack, the initial parameters of crack extension are set, including the initial crack length, crack extension direction, and loading conditions.

9. The method for predicting fatigue crack growth life of FSW joints based on PSO-SVR according to claim 1, characterized in that: The fatigue crack growth is jointly extended by the crack growth rate model and the crack tip stress intensity factor prediction model: In the formula, a k is the crack length under the kth crack extension step; Δa is the crack extension growth step length; N k is the fatigue life under the kth crack extension step; for a k-1 The crack growth rate under .

10. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory. When the computer program is executed by the processor, the method for predicting fatigue crack growth life of a FSW joint based on PSO-SVR as described in any one of claims 1 to 9 is implemented.

Images