Multi-modal data driven metal multi-axial fatigue life prediction method

Abstract

The invention discloses a multi-modal data-driven metal multi-axial fatigue life prediction method, which takes multi-modal data as a data driving basis, provides a new fatigue life prediction framework by utilizing a fusion machine learning model, and can be used as an alternative model for fatigue life prediction. The method comprises the following steps: firstly, defining multi-precision data: experimental data obtained through an actual experiment and simulation data obtained through physical simulation; the mechanical properties of the material are obtained through preliminary experiments, and a reliable finite element analysis (FEA) model is established according to the properties. Then, a high-reliability data set is generated by using a finite element analysis model. In order to process data with multi-modal features, the method comprises the step of constructing a hybrid prediction model DL-LGBM-DRS based on dynamic regression selection. The model can learn the characteristics of discrete and time sequence samples at the same time with high accuracy, so that the mixed prediction model DL-LGBM-DRS can efficiently and accurately predict the cyclic fatigue life under various multi-axis loading conditions.

Description

Technical Field

[0001] The present invention belongs to big data analysis and application, and relates to the field of fatigue life prediction based on multimodal physical information, especially low-cycle fatigue life prediction technology under complex multi-axial loads, and more particularly to a multimodal data-driven metal multi-axial fatigue life prediction method. Background Art

[0002] There are two bottleneck problems that need to be solved in the existing prediction methods of fatigue life under multiaxial loads, namely, the physical relationship of fatigue failure under multiaxial loads is complex and difficult to predict; the prediction cost based on statistical methods or experimental methods is high. In the case of multiaxial fatigue where only flat loads (such as proportional or constant amplitude cyclic loads) are considered, the yield criterion can be used to replace the equivalent stress, that is, the equivalent stress-strain criterion. Some scholars have defined multiple equivalent parameters for external phases and non-proportional conditions based on the maximum prism and minimum Frobenius criteria. However, for more complex load conditions, the expected life given by this method is not reliable. Later, it was found that stress along the shear direction would significantly increase the growth rate of fatigue cracks, so the critical surface method was proposed. The critical surface method evaluates multiaxial fatigue based on damage parameters calculated from stress and / or strain components, and has become a popular method due to its effectiveness in multiaxial fatigue problems. The critical surface method can also determine the location of fatigue cracks under various cyclic fatigue states.

[0003] With the development of technology, the prior art discloses: fatigue analysis under multiaxial load using finite element method with built-in physical criteria. Finite element analysis (FEA) is a method based on physical laws that is widely used in fatigue analysis. The finite element analysis method is able to simulate fatigue conditions under compiled loading spectra, such as stress-strain distribution under full multiaxial non-proportional loading. The fatigue state of the surface can also be well described based on roughness and geometric properties. In addition to mechanical analysis, many studies also focus on crack propagation. The extended finite element method (XFEA) based on fracture mechanics can reveal the crack propagation trend of metal parts. In addition, using the stress-strain information output by finite element analysis, fatigue analysis software such as Fe-Safe can be used to directly perform life assessment. In order to improve efficiency and reliability, researchers have tried to develop alternative methods while avoiding a large number of nonlinear operations in finite element analysis. Considering that uncertain factors (such as material properties, geometry, and loads) can lead to dispersion of fatigue life, statistical methods have been proposed to obtain a more reasonable fatigue life distribution. In the case of probabilistic fatigue life prediction, Monte Carlo simulation (MCS) is often used as a general prediction framework. For multivariate uncertainty assessment, the processing of a large number of input influencing parameters and output results is challenging. MCS can solve this problem by combining the weighting law in structural reliability.

[0004] Although statistical knowledge-based methods can predict fatigue life with high accuracy and efficiency, there are still some limitations in the implementation process and prediction results: (i) it is difficult to obtain a sufficient amount of statistical data; (ii) the interpretability and fidelity are low compared with physics-based methods; (iii) the computational cost is expensive when the uncertainty is considerable. Therefore, the growing demand for fatigue prediction requires novel hybrid methods. Physically informed machine learning (PIML) is an emerging paradigm that incorporates physical principles as "prior" knowledge to significantly improve the performance of machine learning models. One approach to PIML is to embed physical or empirical laws (such as ordinary differential equations) into machine learning networks. Therefore, in practical engineering applications, insufficient data dimensionality and quantity will seriously affect the training performance of PIML. As mentioned above, finite element analysis can be used as an appropriate alternative to data acquisition based on physical laws, which is common in many engineering projects. It has been proven that PIML based on finite element analysis is a very reliable and efficient method in the absence of data.

[0005] Multiaxial fatigue life prediction is very complex and involves many factors, especially input factors with multimodal characteristics. The basic physical properties of the material determine the fatigue behavior and change process of the material under complex external loads. The external loads are time-series and will iteratively affect the structural response through geometric effects. A large number of studies have shown that physical response and fatigue life are strongly correlated, but there is a lack of methods for prediction using both discrete and time-series characteristics.

[0006] In summary, the existing technology mainly has the following deficiencies:

[0007] 1. Data-driven methods lack physical guidance;

[0008] 2. Data-driven methods lack data sources in complex load fields;

[0009] 3. Fatigue life prediction is costly, slow and inaccurate. Summary of the invention

[0010] Purpose of the invention: In view of the above-mentioned deficiencies in the prior art, the present invention provides a multi-modal data driven metal multi-axial fatigue life prediction method.

[0011] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical solution.

[0012] A multi-modal data driven metal multi-axial fatigue life prediction method comprises the following steps:

[0013] S1. Establish a database of properties of aerospace materials to be analyzed based on preliminary experiments, fit engineering equations through experimental data, obtain the physical properties of the material, and add these properties into the data set as data with discrete characteristic properties;

[0014] S2. Establish a finite element model based on the physical properties of the material. By combining finite element analysis with actual test comparison, determine the fatigue life as a label to establish a high-reliability training data set based on simulation. This training data set is used for machine learning algorithms to train the learning of physical relationships.

[0015] S3, the data set includes multi-axial load spectra with time series characteristics, time series physical responses, and materials and external environment information with discrete characteristics;

[0016] This step includes constructing a hybrid prediction algorithm DL-LGBM-DRS based on dynamic regression selection, thereby realizing the prediction of multiaxial fatigue life of metal materials under multi-modal input characteristics. The algorithm establishment steps include:

[0017] S31. By introducing the gradient-based unilateral sampling method and the mutually exclusive feature merging method, the LightGBM model is constructed, and then the GBDT algorithm is used to combine multiple decision trees to process dense discrete data, achieving high interpretability and generalization.

[0018] S32. Constructing a deep learning model based on a long short-term memory network based on a recurrent neural network, wherein the long short-term memory network is constructed through a forget gate, an input gate, and an output gate to realize a long-term memory function;

[0019] S33. Build a dynamic regression selection algorithm to fuse the LightGBM model and the deep learning model to form a fatigue life prediction algorithm that processes multimodal inputs;

[0020] The dynamic regression network is as follows:

[0021] Input set F, training set T, test set S, weight vector W, define the neighbor as K, and the independent model number as n;

[0022] Output First define MAE as mean absolute error:

[0023]

[0024] In the formula, y is the actual value of the training sample, is the predicted value of the sample;

[0025] Initialize the mean absolute error, define error Error = 0 and the empty set A, then for each sample x in the test set S i, find the competition interval ψ of the sample corresponding to the training set T, and the value predicted by each independent model in each set group F Calculate the mean absolute error about the competition interval ψ. The calculation process includes:

[0026] Definition n =w×MAE n , then A=A∪a n , until all model prediction results and error means are:

[0027]

[0028] and Represents the update of the set, for each predicted value in the set group If a n >(a max -a min ) / 2, then the model selection is performed and the following updates are performed:

[0029]

[0030] Define N as the size of the cluster mean, for each an Perform model combination, namely:

[0031]

[0032] Now define:

[0033]

[0034] The final error is Error fin =Error / |X|;

[0035] The dynamic regression selection algorithm finally evaluates the weight of each model, which is defined by the final error Errorfin above, f(x j ) is the observed value of the corresponding sample in the test set, is the normalized result;

[0036] For each sample, the value with the smaller error is judged as the optimal one, and the local optimal prediction result is the final prediction result of the dynamic regression network algorithm.

[0037] Furthermore, the steps of establishing the hybrid prediction model DL-LGBM-DRS for dynamic regression selection in step S3 include:

[0038] S31. Use the GBDT algorithm to combine multiple decision trees to process dense discrete data to achieve high interpretability and generalization. The calculation is as follows:

[0039] The standard decision tree is as follows:

[0040]

[0041] In the formula, X1, X2 are the training subsets of each branch, yi is the actual value of the sample, is the branch prediction value of the sample;

[0042] The GBDT algorithm requires the following sub-steps:

[0043] 1) Initialize the weak learner and define the loss function L(y,f(x)) and the prediction function g0(x):

[0044]

[0045] In the formula, C m Represented as the mean of sample y in m trees;

[0046] 2) For M iterations, the residual is estimated as:

[0047]

[0048] The regression tree then starts from {(x i ,r im ), i = 1...n}, and m trees and J leaf nodes are R mj (j=1,2,...,J), and then cmj Find the optimal value and minimize the loss function at the same time, that is:

[0049]

[0050] At this time c mj For R mj The minimum square loss is:

[0051]

[0052] I is the indicator function, if x∈R mj Then I = 1; otherwise I = 0;

[0053] 3) The strong learner will be updated as:

[0054]

[0055] 4) The iteration stops when the decision function converges, and the final model with higher accuracy is:

[0056]

[0057] After that, the LightGBM model was constructed by introducing the gradient-based single-side sampling method (GOSS) and the mutually exclusive feature merging (EFB) method.

[0058] Furthermore, step S1 includes using quasi-static uniaxial tensile tests, uniaxial tension and compression tests, pure torsion tests and multiaxial fatigue tests to make a preliminary estimate of the physical properties of the required aviation materials, fitting the test data using the Manson-Coffin equation and the Ramberg-Osgood equation, and then extracting the equation characteristics to obtain the uniaxial and cyclic fatigue properties of the material.

[0059] Furthermore, the combined finite element analysis method described in step S2 specifically includes the following operations:

[0060] In the physical response calculation stage, the finite element calculation solver ABAQUS / Standard is used with an embedded mixed cycle model to achieve the physical calculation effect of the Chaboche mixed cycle model. At the same time, the Ramberg-Osgood coefficient is used for plastic correction to ensure the accuracy of the physical response calculation under multi-axial loads.

[0061] During the fatigue life calculation stage, the finite element solver Simulia / Fe-Safe is used, and the Brown-Miller multi-axial load fatigue life estimation method is adopted. The load spectrum and physical response are input to calculate the fatigue life. Finally, three physical response quantities including displacement, stress and strain are obtained. These data with time series characteristics will be used to establish a data set.

[0062] Furthermore, the data with discrete properties described in step S1 includes:

[0063] Basic properties: including material density and elastic modulus;

[0064] Hardening model parameters: including yield limit, tensile limit, Poisson's ratio, elongation and section reduction;

[0065] Cyclic fatigue parameters: including uniaxial tensile and compressive strain amplitude, uniaxial tensile and compressive stress amplitude, cyclic hardening coefficient and cyclic hardening exponent.

[0066] Beneficial effects: Compared with the prior art, the multi-modal data-driven metal multi-axial fatigue life prediction method provided by the present invention can establish a highly reliable data set through physical knowledge guidance and finite element method when the training data set is insufficient. At the same time, multi-modal data input can be used to predict the fatigue life under multi-axial load with higher speed and accuracy, which is beneficial to practical engineering applications, especially in application scenarios with higher requirements for real-time prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1It is a schematic diagram of the framework flow of the method of the present invention;

[0068] Figure 2 It is a network structure diagram of the dynamic regression classification hybrid prediction method proposed by the present invention;

[0069] Figure 3 It is a fitting schematic diagram of a preliminary experimental embodiment of the present invention;

[0070] Figure 4 It is the algorithm effect diagram in the embodiment;

[0071] Figure 5 It is an effect diagram obtained by implementing the present invention in predicting fatigue life of other aviation materials. DETAILED DESCRIPTION

[0072] In order to explain the technical solution disclosed by the present invention in detail, a specific introduction is given below in conjunction with the accompanying drawings of the specification.

[0073] The present invention belongs to big data methods and applications, and specifically relates to the field of fatigue life prediction based on physical information, especially low-cycle fatigue life prediction technology under complex multi-axial loads, and in particular a multi-modal data-driven metal multi-axial fatigue life prediction method. The fatigue life prediction framework constructed by the present invention uses a physics-guided machine learning method as an alternative model for fatigue life prediction, which can predict fatigue life under multi-axial loads with higher computational efficiency while ensuring high accuracy.

[0074] Combination Figure 1 The hybrid prediction framework structure shown in the figure, the method described in the present invention adopts a high-reliability data set based on physical information and finite element analysis for training. With the help of test experiments, the required physical properties of aerospace materials within a specific range can be obtained. On this basis, the combined finite element analysis method is used to obtain the physical response of the parts associated with the geometric properties and the low-cycle fatigue life under the corresponding load conditions. However, this method requires a lot of computing cost and time cost. In this regard, the present invention uses ABAQUS / CAE and Simulia Fe-Safe solver to perform finite element solution to obtain a high-reliability data set associated with fatigue life for subsequent training of the agent model based on machine learning.

[0075] Considering that LightGBM, as an integrated learning model, has a more efficient computing speed than other base models and is very flexible in dealing with discrete data, its modeling ability for time series data is limited. The deep learning method of long short-term memory network is good at time series and dependencies, and can automatically perform feature engineering, which has the disadvantage of high requirements for computing resources. Therefore, the present invention designs a fusion prediction strategy that can combine the advantages of both, fully utilize the input features, and improve the computing efficiency and prediction accuracy.

[0076] For fusion, the present invention proposes a dynamic regression selection algorithm based on the K-nearest neighbor algorithm, which can select a regression model suitable for dynamic system modeling. It combines regression analysis with time series prediction and adapts to data changes by dynamically adjusting the model.

[0077] Finally, the LightGBM model, long short-term memory network and dynamic regression selection algorithm are combined to train the two models on the previous data set, and the mean absolute error is used as an indicator to select the optimal model for each prediction sample to obtain the final prediction result.

[0078] Specifically, the present invention provides a multi-modal data driven metal multi-axial fatigue life prediction method comprising the following steps:

[0079] (1) For metal materials, especially multiphase metals used in complex environments such as aerospace and deep-sea exploration, their physical properties are relatively complex. More importantly, it will cost a lot to determine the distribution of their physical responses based on actual experiments or to conduct fatigue life experiments on them. In order to obtain highly reliable results in the future, the present invention first establishes a property database of aerospace materials to be analyzed based on preliminary experiments, and obtains the physical properties of the materials by fitting engineering equations through experimental data.

[0080] This step includes using quasi-static uniaxial tensile tests, uniaxial tension and compression tests, pure torsion tests and multiaxial fatigue tests to make a preliminary estimate of the physical properties of the required aviation materials, fitting the test data using the Manson-Coffin equation and the Ramberg-Osgood equation, and then extracting the equation characteristics to obtain the uniaxial and cyclic fatigue properties of the material.

[0081] As attached Figure 2 As shown in FIG. 1 , an example of material property fitting for TC4 alloy is shown. The relationship between the Manson-Coffin equation and the Ramberg-Osgood equation has a good correlation with the test results. Among them, Figure 2 (a) is the fitting result of the Ramberg-Osgood equation, through which the stress-strain relationship of the experimental TC4 alloy under the uniaxial state is obtained. Figure 2 (b) is the Manson-Coffin equation, which gives the relationship between the material strain and low-cycle fatigue life. Through these, the fatigue properties of the material required for simulation are finally obtained.

[0082] (2) After obtaining the basic physical properties, for the subsequent data-driven requirements, it is necessary to establish a finite element model that can provide a data set. The present invention establishes a finite element model based on the physical properties of the material, and determines the fatigue life as a label to establish a training data set by combining finite element analysis with actual test comparison. The data set is used for hidden relationship training of the machine learning algorithm;

[0083] This step builds a high-accuracy finite element model based on physical properties, uses combined finite element analysis, compares it with actual tests, and uses deterministic fatigue life as a label to build a highly reliable training data set.

[0084] The combined finite element analysis method includes the following operations:

[0085] In the physical response calculation stage, the finite element solver ABAQUS / Standard is used with an embedded mixed cycle model to achieve the physical calculation effect of the Chaboche mixed cycle model. At the same time, the Ramberg-Osgood coefficient is used for plastic correction. This method can ensure the accuracy of the physical response calculation under multi-axial loads.

[0086] In the fatigue life calculation stage, the finite element calculation solver Simulia / Fe-Safe is used, and the Brown-Miller multi-axial load fatigue life estimation method is adopted to input the load spectrum and physical response to calculate the fatigue life.

[0087] (3) Fatigue prediction under multi-axial load conditions is a very complex problem, one of which is that the input information has multimodal characteristics. It can be simply classified as load spectrum input information with time series characteristics, and it also belongs to external environment information in fatigue life prediction. In addition, it includes discrete physical response information generated after the material is subjected to external loading. In order to better train the data set containing the above input features obtained by combining the finite element method in the previous step, the present invention proposes a hybrid prediction model DL-LGBM-DRS based on deep learning and LightGBM algorithm to construct dynamic regression selection, thereby realizing the prediction of low-cycle fatigue life under multi-modal input environment.

[0088] In summary, this step is based on deep learning and LightGBM model, and proposes a hybrid prediction model DL-LGBM-DRS with dynamic regression selection. The network structure is shown in the attached figure. Figure 3 As shown, the prediction of low cycle fatigue life under multi-modal input environment is achieved.

[0089] The steps of establishing the hybrid prediction model DL-LGBM-DRS with dynamic regression selection include:

[0090] (3.1) The GBDT algorithm is used to combine multiple decision trees to process dense discrete data and achieve high interpretability and generalization. The calculation is as follows:

[0091] The standard decision tree is as follows:

[0092]

[0093] In the formula, X1 and X2 are the training subsets of each branch respectively. yi , is the actual value of the sample, is the branch prediction value of the sample.

[0094] The GBDT algorithm requires the following sub-steps:

[0095] (3.1.1) Initialize the weak learner and define the loss function L(y,f(x)) and prediction function g0(x):

[0096]

[0097] In the formula, C m Represented as the mean of the samples y in the m trees

[0098] (3.1.2) For M iterations, the residual estimate is:

[0099]

[0100] The regression tree then starts from {(x i ,r im ), i = 1...n}, and m trees and J leaf nodes are R mj (j=1,2,...,J), and then cmj Find the optimal value and minimize the loss function at the same time, that is:

[0101]

[0102] At this time c m j is R m The minimum square loss of j is:

[0103]

[0104] I is the indicator function, if x∈R mj Then I = 1; otherwise I = 0.

[0105] (3.1.3) The strong learner will be updated as:

[0106]

[0107] (3.1.4) The iteration will stop when the decision function converges, and the final model with higher accuracy is:

[0108]

[0109] After that, the LightGBM model was constructed by introducing the gradient-based single-side sampling method (GOSS) and the mutually exclusive feature merging (EFB) method.

[0110] (3.2) According to the recurrent neural network (RNN) model, the present invention constructs a deep learning model based on the long short-term memory network, wherein the long short-term memory network is constructed by a forget gate, an input gate, and an output gate to realize a long-term memory function.

[0111] (3.3) Using the dynamic regression selection method, the LightGBM model and the deep learning model are integrated to obtain a low-cycle fatigue life prediction algorithm with multi-modal input. The specific steps of constructing the dynamic regression network algorithm include:

[0112] S33. Design a dynamic regression selection algorithm that can effectively integrate the LightGBM model and the deep learning model to form a fatigue life prediction algorithm that processes multimodal inputs;

[0113] The dynamic regression network is as follows:

[0114] Input set F, training set T, test set S, weight vector W, define the neighbor as K, and the independent model number as n;

[0115] Output First define MAE as mean absolute error:

[0116]

[0117] In the formula, y is the actual value of the training sample, is the predicted value of the sample;

[0118] Initialize the mean absolute error, define error Error = 0 and the empty set A, then for each sample x in the test set S i , find the competition interval ψ of the sample corresponding to the training set T. For each independent model prediction value in each set group F Compute the mean absolute error with respect to the contention interval ψ:

[0119] Definition n =w×MAE n , then A=A∪a n ;

[0120] Until all model prediction results and error means are:

[0121]

[0122] For each predicted value in the cluster If a n >(a max -a min ) / 2 Then the model selection is performed, at this time:

[0123]

[0124] Define N as the size of the cluster mean, for each a n Perform model combination, namely:

[0125]

[0126] Now define:

[0127]

[0128] The final error is

[0129] Error fin =Error / |X|

[0130] The dynamic regression selection algorithm finally evaluates the weight of each model, which is defined by the final error Errorfin above. f(xj) is the observed value of the corresponding sample in the test set, For the two models used in the present invention, for each sample, the value with the smaller error is determined to be the best, and the local best prediction result is the final prediction result of the dynamic regression network algorithm.

[0131] Example

[0132] First, the method described in the present invention was experimentally applied to the fatigue test data set of shaft samples to verify the effectiveness of the proposed model. The models selected for comparison were LSTM network without inputting load spectrum (DL w / o loadingspecturms), independent deep learning network, independent LightGBM, XGBoost and random forest. The evaluation indicators included root mean square error (RMSE), mean absolute error (MAE), regression coefficient (R 2 ) and calculation speed. The experimental results are respectively in the appendix of the manual. Figure 4 Displayed in.

[0133] The fusion model proposed in the present invention has better performance in predicting simulated fatigue life. The prediction performance of LGBM is good, showing its high learning efficiency on discretized feature data, and it is a potential fatigue behavior prediction model. The results of deep learning are worthy of attention. The accuracy of the deep learning model using the fusion of loading spectrum and physical response as input is slightly lower than that of the proposed DRS model. However, after removing the loading spectrum, the accuracy of deep learning dropped significantly, while the error increased significantly. This difference highlights the importance of load spectrum information as input. In addition, we also compared other GBDT-based algorithms, including XGBoost and random forest, and their gradual decline in prediction performance is consistent with the pattern we found in previous studies. In general, the errors of machine learning models are within an acceptable range, and the prediction accuracy of fusion models is higher than that of single models. In terms of time cost, due to the addition of a decision layer that plays a role in late fusion, the fusion model is slower than the independent model. However, compared with the finite element analysis method, the machine learning model used for environmental applications in this study has an incomparable speed advantage. Appendix Figure 4 c shows the comparison between the fatigue life predicted by the machine learning model and the fatigue life calculated by the finite element analysis method. It is clear that the proposed hybrid prediction DL-LGBM-DRS model shows very favorable results compared to other models when predicting fatigue life based on the generated dataset. This proves the feasibility of using the DL-LGBM-DRS model as an alternative model for fatigue life prediction, while maintaining high accuracy and improving efficiency.

[0134] Subsequently, the method described in the present invention was tested on other commonly used metal materials in aerospace to verify the effectiveness of the model in predicting fatigue life under multi-axial loads. The metals selected for comparison are 304 stainless steel, pure titanium, GH4169 and 7075-T651 aluminum alloys. The prediction results are shown in the attached manual. Figure 5 shown.

[0135] Figure 5 (a) shows the prediction results of 304 stainless steel. Figure 5 (b) is the prediction result of 7075-T651. Figure 5 (c) shows the prediction result of pure titanium, Figure 5(d) is the prediction result of GH4169 superalloy. The prediction sample is consistent with TC4 alloy, and the dotted lines are 1.5 times and two times scatter bands to verify the accuracy of the prediction results. The results show that the fitting results of 304 stainless steel, pure titanium and GH4169 superalloy are quite good; the fitting points are located in the two scatter bands. In the case of GH4169, it can be observed that most of the fitting points are below the fitting line, which indicates that the model is pessimistic about the fatigue life prediction. This phenomenon is consistent with the fatigue life trend calculated using the Brown-Miller method. The prediction results of 7075-T651 aluminum alloy case deviate from the experimental results. Due to cost constraints, we conducted fewer experiments in the case of 7075-T651 alloy. This may lead to a certain probability distribution of experimental fatigue life. In addition, 7075-T651 alloy shows higher fatigue life under comparable loading conditions due to its material properties. In this case, the fusion method guided by the physical model is not sufficient to predict high cycle fatigue life. Since the mechanical properties of the materials were introduced differently in this experiment to save costs, there are slight differences between the material properties, which lead to differences in the final prediction results. In summary, the proposed method is satisfactory in predicting low-cycle fatigue life under multiaxial loading. In addition, in the case of different alloys, the results obtained by the proposed framework are in good agreement with the experimental fatigue life. This reveals the potential ability of the proposed method to be used for fatigue life prediction of other metallic materials.

Claims

1. A multi-modal data driven metal multiaxial fatigue life prediction method, characterized in that: The steps include: S1. Establish a database of properties of aerospace materials to be analyzed based on preliminary experiments, fit engineering equations through experimental data, obtain the physical properties of the material, and add these properties into the data set as data with discrete characteristic properties; S2. Establish a finite element model based on the physical properties of the material. By combining finite element analysis with actual test comparison, determine the fatigue life as a label to establish a high-reliability training data set based on simulation. This training data set is used for machine learning algorithms to train the learning of physical relationships. S3, the data set includes multi-axial load spectra with time series characteristics, time series physical responses, and materials and external environment information with discrete characteristics; This step includes constructing a hybrid prediction algorithm DL-LGBM-DRS based on dynamic regression selection, thereby realizing the prediction of multiaxial fatigue life of metal materials under multi-modal input characteristics. The algorithm establishment steps include: S31. By introducing the gradient-based unilateral sampling method and the mutually exclusive feature merging method, the LightGBM model is constructed, and then the GBDT algorithm is used to combine multiple decision trees to process dense discrete data, achieving high interpretability and generalization. S32. Constructing a deep learning model based on a long short-term memory network based on a recurrent neural network, wherein the long short-term memory network is constructed through a forget gate, an input gate, and an output gate to realize a long-term memory function; S33. Build a dynamic regression selection algorithm to fuse the LightGBM model and the deep learning model to form a fatigue life prediction algorithm that processes multimodal inputs; The dynamic regression network is as follows: Input set F, training set T, test set S, weight vector w, and independent model number n; Output First define MAE as mean absolute error: In the formula, y is the actual value of the training sample, is the predicted value of the sample; Initialize the mean absolute error, define error Error = 0 and the empty set A, then for each sample x in the test set S i , corresponding to the training set T to find the competition interval of the sample ψ , for each independent model prediction value in each cluster F Calculate the contention interval ψ The mean absolute error is calculated as follows: Definition n =w×MAE n , then A=A∪a n , until all model prediction results and error means are: and Represents the update of the set, for each predicted value in the set group If a n >(a max -a min ) / 2, then the model selection is performed and the following updates are performed: Define N as the size of the cluster mean, for each an Perform model combination, namely: Now define: The final error is Error fin =Error / |X|; The dynamic regression selection algorithm finally evaluates the weight of each model, which is defined by the final error Errorfin above, f(x j ) is the observed value of the corresponding sample in the test set, is the normalized result; For each sample, the value with the smaller error is judged as the optimal one, and the local optimal prediction result is the final prediction result of the dynamic regression network algorithm.

2. The multi-modal data driven metal multi-axial fatigue life prediction method according to claim 1 is characterized in that: The calculation steps of the GBDT algorithm for processing the discrete part of the input in step S31 include: The GBDT algorithm is used to combine multiple decision trees to process dense discrete data and achieve high interpretability and generalization. The calculation is as follows: The standard decision tree is as follows: In the formula, X1, X2 are the training subsets y of each branch respectively. i , is the actual value of the sample, is the branch prediction value of the sample; The GBDT algorithm requires the following sub-steps: 1) Initialize the weak learner and define the loss function L(y,f(x)) and the prediction function g0(x): In the formula, C m It is represented as the mean of the samples y in m trees; 2) For M iterations, the residual is estimated as: The regression tree then starts from {(x i ,r im ), i = 1...n}, and m trees and J leaf nodes are R mj (j=1,2,...,J), and then cmj Find the optimal value and minimize the loss function at the same time, that is: At this time c mj For R mj The minimum square loss is: I is the indicator function, if x∈R mj Then I = 1; otherwise I = 0; 3) The strong learner will be updated as: 4) The iteration stops when the decision function converges, and the final model with higher accuracy is: After that, the LightGBM model was constructed by introducing the gradient-based one-side sampling method and the mutually exclusive feature merging method.

3. The multi-modal data driven metal multi-axial fatigue life prediction method according to claim 1, characterized in that: Step S1 includes using quasi-static uniaxial tensile test, uniaxial tension and compression test, pure torsion test and multiaxial fatigue test to make a preliminary estimate of the physical properties of the required aviation materials, fitting the test data using the Manson-Coffin equation and the Ramberg-Osgood equation, and then extracting the equation characteristics to obtain the uniaxial and cyclic fatigue properties of the material.

4. The multi-modal data driven metal multi-axial fatigue life prediction method according to claim 1, characterized in that: The combined finite element analysis method described in step S2 specifically includes the following operations: In the physical response calculation stage, the finite element calculation solver ABAQUS / Standard is used with an embedded mixed cycle model to achieve the physical calculation effect of the Chaboche mixed cycle model. At the same time, the Ramberg-Osgood coefficient is used for plastic correction to ensure the accuracy of the physical response calculation under multi-axial loads. During the fatigue life calculation stage, the finite element solver Simulia / Fe-Safe is used, and the Brown-Miller multi-axial load fatigue life estimation method is adopted. The load spectrum and physical response are input to calculate the fatigue life. Finally, three physical response quantities including displacement, stress and strain are obtained. These data with time series characteristics will be used to establish a data set.

5. The multi-modal data driven metal multi-axial fatigue life prediction method according to claim 1, characterized in that: The data with discrete properties described in step S1 include: Basic properties: including material density and elastic modulus; Hardening model parameters: including yield limit, tensile limit, Poisson's ratio, elongation and section reduction; Cyclic fatigue parameters: including uniaxial tensile and compressive strain amplitude, uniaxial tensile and compressive stress amplitude, cyclic hardening coefficient and cyclic hardening exponent.

Images