CN113343566A - Deep learning-based nickel-based alloy fracture toughness prediction method and system - Google Patents

Deep learning-based nickel-based alloy fracture toughness prediction method and system Download PDF

Info

Publication number
CN113343566A
CN113343566A CN202110601155.0A CN202110601155A CN113343566A CN 113343566 A CN113343566 A CN 113343566A CN 202110601155 A CN202110601155 A CN 202110601155A CN 113343566 A CN113343566 A CN 113343566A
Authority
CN
China
Prior art keywords
fracture toughness
nickel
feature vector
deep learning
experimental data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110601155.0A
Other languages
Chinese (zh)
Other versions
CN113343566B (en
Inventor
徐雅斌
崔露露
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Information Science and Technology University
Original Assignee
Beijing Information Science and Technology University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Information Science and Technology University filed Critical Beijing Information Science and Technology University
Priority to CN202110601155.0A priority Critical patent/CN113343566B/en
Publication of CN113343566A publication Critical patent/CN113343566A/en
Application granted granted Critical
Publication of CN113343566B publication Critical patent/CN113343566B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/26—Composites
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Abstract

The invention relates to a nickel-based alloy fracture toughness prediction method and system based on deep learning, which are characterized in that a deep confidence network and a support vector regression network based on an attention mechanism are combined together to predict fracture toughness values, and the strong feature extraction capability of a deep learning model and the regression prediction capability of support vector regression are organically combined. Meanwhile, feature vectors obtained by a depth confidence network based on an attention mechanism and original features are spliced and input into a support vector regression model, so that the fracture toughness prediction method provided by the invention simultaneously considers high-order combination features and low-order linear features related to fracture toughness and has higher accuracy.

Description

Deep learning-based nickel-based alloy fracture toughness prediction method and system
Technical Field
The invention relates to the technical field of toughness prediction, in particular to a nickel-based alloy fracture toughness prediction method and system based on deep learning.
Background
With the continuous development of the aerospace industry, the requirements on the performance and reliability of an aero-engine are continuously improved, and the requirements on the comprehensive performance, the temperature bearing capacity and the like of materials are higher and higher. The nickel-based superalloy still has high strength and excellent oxidation resistance and corrosion resistance under high temperature conditions, so that the nickel-based superalloy becomes a preferred material for key hot-end components of modern aeroengines, spacecrafts, rocket engines and the like.
Fracture toughness is a toughness parameter of a material for resisting brittle failure, is a critical value for judging whether a material crack reaches an unstable level, and plays an important role in damage tolerance design and structural integrity evaluation. For parts working at high temperature, the temperature distribution is often in large gradient, and the fracture toughness measured at room temperature is far from meeting the design requirements of actual engineering. Therefore, there is a strong need for a method for accurately predicting fracture toughness in nickel-base superalloys.
Disclosure of Invention
The invention aims to provide a nickel-based alloy fracture toughness prediction method and system based on deep learning, which can accurately predict the fracture toughness of nickel-based high-temperature alloys with different temperatures and different components according to the existing experimental data.
In order to achieve the purpose, the invention provides the following scheme:
a method for predicting fracture toughness of a nickel-based alloy based on deep learning, the method comprising:
preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
learning and optimizing features of the experimental data set by using a depth confidence network based on an attention mechanism to obtain an optimized feature vector;
splicing the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
and predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted feature vector.
The invention also provides a nickel-based alloy fracture toughness prediction system based on deep learning, which comprises:
the input module is used for preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
the deep learning module based on attention is used for learning and optimizing the features of the experimental data set by using a deep confidence network based on an attention mechanism to obtain an optimized feature vector;
the feature splicing module is used for splicing the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
and the output module is used for predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted characteristic vector.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides a nickel-based alloy fracture toughness prediction method and system based on deep learning, wherein a deep confidence network based on an attention mechanism and a support vector regression model are combined together to predict a fracture toughness value, and the strong feature extraction capability of the deep learning model and the regression prediction capability of the support vector regression are organically combined. Meanwhile, feature vectors obtained by a depth confidence network based on an attention mechanism are spliced with original features and input into a support vector regression model, so that the prediction method provided by the invention considers high-order combination features and low-order linear features related to the fracture toughness of the nickel-based alloy, and has higher accuracy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a flowchart of a method for predicting fracture toughness of a nickel-based alloy based on deep learning according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a depth confidence network structure based on an attention mechanism according to an embodiment of the present invention;
fig. 3 is a block diagram of a system for predicting fracture toughness of a nickel-based alloy based on deep learning according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The fracture toughness of a superalloy strongly depends on alloying elements and experimental conditions, but is difficult to characterize by a specific mathematical expression due to its complex and highly nonlinear relationship. Thus, the variety and interplay of alloying elements tends to make it difficult to determine an appropriate fracture toughness. In solving this problem, materials genetic engineering techniques have shown great potential.
The material genetic engineering is a subversive leading-edge technology in the international material field in recent years, and the basic idea is to combine the high-throughput calculation, high-throughput experiment and material big data technology of the material, accelerate the research and development process of the material from discovery, manufacture to application and reduce the cost through synergistic innovation. The data and artificial intelligence are the core of material genetic engineering, and through the application of big data and artificial intelligence technology, the problems of correlation and process optimization between the new material organization structure-performance-process can be realized, and the performance of the material is improved.
The invention aims to provide a nickel-based alloy fracture toughness prediction method and system based on deep learning, which can reduce the experimental amount and the experimental cost based on the concept and method of material genetic engineering, and have great theoretical significance and application value for promoting the research and development of the nickel-based superalloy technology.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example 1
The direct method for obtaining the relationship between the fracture toughness and the temperature of the high-temperature alloy is to perform fracture toughness measurement tests on a standard sample at different temperatures and then use an empirical model for fitting. However, this empirical model is not widely applicable and can only be used for the materials under study. In addition, more data at high temperature is required to improve the accuracy of the model, which causes a great consumption of time and cost.
In order to overcome the limitation of fitting methods based on empirical data, some researchers propose a main curve method deduced based on physical significance, but the method is only suitable for the ductile-brittle transition temperature range, for ferrite steel, the temperature range is usually lower than 0 ℃, and obviously the method is not suitable for nickel-based high-temperature alloys working at high temperature.
A temperature-dependent model has also been proposed to predict the fracture toughness value of the superalloy at high temperatures based on the stress-strain curve and the linear expansion coefficient α of the superalloy, but the model needs to be based on other mechanical property data, and the accuracy of the predicted data is not high.
In summary, the empirical physical model not only requires a large amount of time and cost, but also the accuracy of the predicted data is not high.
Machine learning models have been widely used in recent years to predict mechanical properties of materials. The feedforward neural network is adopted to predict the mechanical properties (ultimate strength, yield strength, strain hardening index and elongation) of the nickel-base superalloy IN 718. One establishes an average static fracture energy prediction model of the nickel-based superalloy IN738LC based on experimental data by utilizing a three-layer feedforward neural network, and reduces the time required by experiments.
The relation between the fracture toughness of the 7075-T651 aluminum alloy and the crack geometry, the load type and the working temperature is analyzed by using a multilayer feedforward sensing artificial neural network, or the quantitative relation between the microstructure and the fracture toughness of the niobium-silicon alloy is established by using a three-layer BP neural network, and the alloy process parameters are optimized by using an artificial neural network model. A prediction model of the mechanical properties (tensile strength, yield strength, elongation, reduction of area and fracture toughness) of the Ti-10V-2Fe-3Al alloy under the conditions of hot working process parameters such as deformation temperature, deformation degree, solid solution temperature, aging temperature and the like and a heat treatment system is built by adopting a 3-layer BP neural network. And a fracture toughness prediction model under different alloy components and low alloy yield stress is established by using a BP neural network with double hidden layers, so that the possibility of designing a new alloy with higher fracture toughness by modifying the chemical components of the alloy is provided.
The fracture toughness of the pearlite steel can be predicted by adopting a Generalized Regression Neural Network (GRNN) optimized based on a fruit Fly Optimization Algorithm (FOA) so as to improve the prediction precision of a small amount of samples, and the influence of alloy elements on the fracture toughness is further researched on the basis of the established model so as to optimize the composition parameters of the steel. The artificial neural network is used by the scholars to establish the relationship between the components and processing parameters and the mechanical properties of the steel plate (yield strength, tensile strength, elongation and ductility) to find out the importance of different variables. Some researchers used 5 machine learning methods to predict four mechanical properties (fatigue strength, tensile strength, breaking strength and hardness) of steel selected from the database of the national institute of materials science, japan.
ANN and ANFIS were also used to predict the fracture energy of polymer nanocomposites and compared to Huang and Kinloch models and three linear regression models. Experimental results show that the prediction effect of ANN and ANFIS is better, and the ANFIS method is the best. Some people adopt a 3-layer BP neural network to predict the fracture toughness of the composite ceramic material. The influence of the composition, the form, the size and the volume fraction of a single phase on the fracture toughness of the Nb-silicide in-situ composite material is researched by using a three-layer feedforward neural network, and the optimal microstructure capable of improving the fracture toughness is predicted by using the model.
However, the shallow learning model based on machine learning is adopted in the above methods, and has a better prediction effect compared with an equation or a model based on experience, but the shallow learning model needs to rely on prior knowledge in a specific field to extract features, and cannot completely reflect essential features of data.
Therefore, the deep learning idea is combined with the support vector regression method, the traffic flow prediction method based on the deep learning regression machine is provided, the data input of influencing factors and the learning and conversion of layer-by-layer information are realized by constructing a multilayer limited Boltzmann machine structure, and equivalent key information is extracted. And then inputting the converted equivalent key information into a support vector regression model to realize short-time traffic flow prediction. The method is characterized in that a deep confidence network model is used for learning the characteristics of traffic flow data in a road network, so that the essential characteristics of the data are mined, and then the support vector regression method is used for predicting the traffic flow. There is also a model combining gate-controlled cycle unit network (GRU) and support vector regression for predicting short-term traffic flow, which uses GRU to extract features of data, and inputs them into the support vector regression model at the top of the model to obtain the final prediction output of the model.
Compared with a single support vector regression model, the prediction model combining deep learning and support vector regression has the advantages that the deep learning model can be used for extracting the features of data, so that the dependence of feature extraction on manpower is reduced, the essential features among data are carved, and the prediction accuracy of the model is improved. In addition, compared with a single deep neural network model, the combined model can utilize the nonlinear kernel function of the support vector regression model to realize a larger generalization performance on data, so that the generalization capability of the model is improved.
Specifically, as shown in fig. 1, the present embodiment provides a method for predicting fracture toughness of a nickel-based alloy based on deep learning, where the method includes:
step 101: preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
step 102: learning and optimizing features of the experimental data set by using a depth confidence network based on an attention mechanism to obtain an optimized feature vector;
step 103: splicing the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
step 104: and predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted feature vector.
In step 101, firstly, 877 pieces of experimental data about the fracture toughness of the nickel-based superalloy are downloaded from a material genetic engineering database; then, the collected data are integrated together according to the sequence of the component contents (C, W, Mo, Nb, Zr, Co, Al, Cr, Ti, Ni), the quenching temperature and the fracture toughness value to form a two-dimensional matrix of (877, 12); then, in order to eliminate the dimensional influence between the indexes, the data is subjected to a most value normalization process, and all the data are mapped between 0 and 1, as shown in the following formula:
Figure BDA0003092982900000061
wherein x isscaleIs the normalized value, x is the original value, xmax、x minRespectively, the maximum and minimum values of the original data set.
Finally, the component contents (C, W, Mo, Nb, Zr, Co, Al, Cr, Ti, Ni) and the quenching temperature of each piece of data are used as input vectors of the model, and the corresponding fracture toughness value is used as an output vector of the model. And according to the ratio of 9: and 1, randomly dividing an experiment training set and an experiment testing set in proportion to obtain an experiment data set.
Then, step 102 is executed: inputting the experimental data set into a deep confidence network to obtain a first feature vector;
inputting the first feature vector into an attention mechanism module to obtain a second feature vector;
and taking the second feature vector as an optimized feature vector.
The structure of the attention-based deep belief network (a-DBN) provided by the present embodiment is shown in fig. 2, wherein the Deep Belief Network (DBN) is a deep neural network composed of two Restricted Boltzmann Machines (RBMs) and one BP neural network unit.
The Restricted Boltzmann Machine (RBM) is a stochastic neural network structure, which is composed of a presentation layer and a hidden layer, wherein the presentation layer is composed of presentation elements and is used as input training data. The hidden layer is composed of hidden elements and serves as a feature detector. When the fracture toughness of the nickel-based superalloy is predicted, firstly, a characteristic matrix is composed of the content and the temperature of each component of the nickel-based superalloy and is input into a display layer of a first RBM network, each layer of RBM network is separately and unsupervised trained, characteristic information is kept as much as possible when characteristic vectors are mapped to different characteristic spaces, and the output of the previous RBM network is used as the input of the next RBM network. And obtaining a first eigenvector X of the fracture toughness of the nickel-based high-temperature alloy after passing through two layers of RBM networks.
Because each layer of RBM network can only ensure that the weight in the layer of RBM network can be optimal for the feature vector mapping of the layer, but not for the feature vector mapping of the whole DBN, a BP neural network is added in the last layer of the DBN, and a network model is trained supervised. In addition, in order to improve the accuracy of model prediction, an attention mechanism module is introduced before the BP neural network. The attention mechanism draws the reference of the processing mode of human vision, focuses attention on a key area, and essentially selects information which plays a key role in a task from a plurality of information, so that the task complexity is reduced, and the accuracy of the model is improved. In the module, a first feature vector X obtained through RBM network learning is input to an attention mechanism layer to obtain a second feature vector a. The second feature vector a is calculated as (1):
Figure BDA0003092982900000071
wherein alpha isiFor attention distribution, the degree of correlation of the ith information in the input information vector X with the query vector q is shown given the query vector q. The calculation formula is shown as (2) and (3):
Figure BDA0003092982900000072
Figure BDA0003092982900000073
wherein, s (x)iAnd q) is an attention scoring function. W, U and V are weight matrixes.
Then, we input the second feature vector a into the BP neural network layer and propagate the error information from top to bottom into the RBM of each layer by using the back propagation network, thereby fine-tuning the whole A-DBN network. Thus, the optimized feature vector F is obtained by utilizing the fine-tuned depth confidence network based on the attention mechanism.
And then splicing the optimized characteristic vector with the original characteristic vector of the experimental data set to obtain a predicted characteristic vector, and inputting the predicted characteristic vector into a support vector regression model to predict the fracture toughness of the nickel-base superalloy.
The support vector regression algorithm is based on VC dimension theory and structure risk minimization principle, can fully use limited sample information, and seeks the optimal balance between the complexity and learning ability of the model to obtain the optimal generalization ability. Conventional regression models typically compute the loss directly based on the difference between the model output f (x) and the true output y, and the loss is zero if and only if the two are identical. In contrast, the support vector regression assumption can tolerate a maximum of ε deviation between the two, i.e., the penalty is only calculated when the absolute value of the difference between f (x) and y is greater than ε. This corresponds to the construction of a spacer band of width 2 epsilon centered on f (x). If the training samples fall within this interval band, the prediction is considered correct. Therefore, the support vector regression problem can be converted to equation (4):
Figure BDA0003092982900000081
c in the formula (4) is a penalty factor lεIs an ε -insensitive loss function, ε is a deviation threshold, W is a weight matrix, f (x)i) As output of the model, yiFor a true output, i is the count variable of the training samples, and m is the total number of training samples.
Support vector regression maps data from a low dimensional space to a high dimensional space through a suitable kernel function, and performs regression in that space to fit a continuous function, minimizing the loss function. There are four common kernel functions, namely a linear kernel function, a polynomial kernel function, a radial basis kernel function and a sigmoid kernel function. Because the radial basis kernel function has the advantages of small parameter variable and selection calculation amount and high calculation efficiency, the radial basis kernel function is selected as the kernel function for supporting vector regression in the embodiment, and the mathematical expression of the radial basis kernel function is as shown in formula (5):
Figure BDA0003092982900000082
in equation (5), σ is the radial basis function width. C and sigma are important parameters related to the generalization performance of the SVR model, and the optimal combination parameters of the C and sigma are optimized and selected by using a grid search algorithm in a K-fold cross validation mode so as to ensure that the established model has optimal performance. The basic principle of the K-fold cross validation is that a data set is divided into K parts with the same size by taking turns K times, the K-1 part is used as a training set, the rest 1 part is used as a validation set, and the average value of the precision of the K-fold validation results is used as an estimated value of modeling precision.
The optimal combination of the penalty factor and the radial basis function width of the support vector regression model is optimally selected by using a grid search algorithm, and the optimal combination comprises the following steps:
listing all possible combinations of penalty factors and radial basis function widths of the support vector regression model, wherein all the possible combinations form a grid;
sequentially carrying out support vector regression modeling on possible combinations in the grid;
and selecting the possible combination with the highest modeling precision as the optimal combination of the penalty factor and the radial basis function width of the support vector regression model by using a k-fold cross verification method.
The time complexity t (n) ═ O (n2) of the algorithm of the prediction method for fracture toughness of nickel-base superalloy based on deep learning provided in this example is mainly derived from the double-layer for loop of lines 1-10, and the codes are as follows:
inputting: c, value of sigma, and a new training set formed by splicing a high-order combination feature vector on the training set obtained by the A-DBN model and the original training set;
and (3) outputting: a support vector regression model constructed by using the optimal parameter combination;
1:for C in[C1,C2…Cm]:
2:forσin[σ1,σ2…σn]:
3, constructing a support vector regression model;
4, performing 10-fold cross validation;
5:if(score>best_score)：
6:best_score＝score；
7, optimal parameter combination { 'C': C, 'sigma': sigma };
8:end if
9:end for
10:end for
11, constructing a support vector regression model by using the optimal parameter combination;
in summary, in the regression prediction process of the fracture toughness of the nickel-based superalloy, firstly, a data set is input into a trained A-DBN model, and a characteristic vector F obtained by the A-DBN model is obtained; then, splicing the feature vector F and the original features to form a new training set and a new testing set, and inputting the new training set and the new testing set into a support vector regression model for training; and finally, predicting the fracture toughness value of the nickel-based superalloy by using the trained support vector regression model.
The method for predicting the fracture toughness of the nickel-based superalloy based on deep learning provided by the embodiment can be used for rapidly and accurately predicting the fracture toughness of the nickel-based superalloy with different temperatures and different components according to the existing experimental data, and effectively solves the problems of complexity, time consumption and low precision in the process of obtaining the fracture toughness value of the nickel-based superalloy by using experiments and an experience-based method.
For the problem of insufficient feature extraction based on artificial experience, in the embodiment, feature extraction is performed by using a deep belief network, compared with the traditional machine learning method that feature selection needs to be performed depending on artificial experience and the problem of essential features of data cannot be completely reflected, the deep belief network has excellent feature learning capability, and the learned features are more essential in data. In addition, in order to optimize the quality of feature extraction, an attention mechanism module is introduced into the deep belief network, different weights are distributed according to the influence degree of each feature on the fracture toughness of the nickel-base superalloy, so that the model can extract features more related to the fracture toughness value of the nickel-base superalloy in the feature learning stage, and the prediction performance of the model is improved.
In order to further improve the regression prediction capability of the model, feature vectors obtained by a depth confidence network module (A-DBN) based on an attention mechanism and original features are spliced and input into a support vector regression model, so that the model considers high-order combination features and low-order linear features at the same time.
Therefore, the method for predicting the fracture toughness of the nickel-based superalloy, which is provided by the embodiment of the invention, can solve the problems of complexity, time consumption and low precision in the process of acquiring the fracture toughness value of the nickel-based superalloy by using an experiment and an experience-based method, can effectively solve the problem that the traditional machine learning method needs to extract features depending on manual experience, optimizes the feature extraction compared with a single deep learning model, considers high-order combination features and low-order linear features at the same time, and improves the prediction precision of the model.
Example 2
A deep learning based nickel-base alloy fracture toughness prediction system, as shown in fig. 3, comprising:
the input module M1 is used for preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
the attention-based deep learning module M2 is used for learning and optimizing the features of the experimental data set by using an attention-based deep belief network to obtain an optimized feature vector;
a feature splicing module M3, configured to splice the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
and the output module M4 is used for predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted characteristic vector.
The emphasis of each embodiment in the present specification is on the difference from the other embodiments, and the same and similar parts among the various embodiments may be referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.

Claims (10)

1. A nickel-based alloy fracture toughness prediction method based on deep learning is characterized by comprising the following steps:
preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
learning and optimizing features of the experimental data set by using a depth confidence network based on an attention mechanism to obtain an optimized feature vector;
splicing the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
and predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted feature vector.
2. The method for predicting fracture toughness of nickel-based alloy based on deep learning as claimed in claim 1, wherein the preprocessing of the experimental data of fracture toughness of nickel-based alloy to obtain the experimental data set comprises:
integrating the fracture toughness experimental data of the nickel-based alloy into a two-dimensional matrix;
and taking the component content and the quenching temperature in the two-dimensional matrix as input, and taking the corresponding fracture toughness value as output to construct an experiment training set and an experiment testing set, wherein the experiment training set and the experiment testing set form an experiment data set.
3. The method for predicting fracture toughness of nickel-based alloy based on deep learning as claimed in claim 1, wherein said learning and optimizing features of said experimental data set by using a deep belief network based on an attention mechanism to obtain an optimized feature vector comprises:
inputting the experimental data set into a deep confidence network to obtain a first feature vector;
inputting the first feature vector into an attention mechanism module to obtain a second feature vector;
and taking the second feature vector as an optimized feature vector.
4. The deep learning-based nickel-base alloy fracture toughness prediction method as claimed in claim 3, wherein the second eigenvector is
Figure FDA0003092982890000011
Wherein alpha isiFor attention distribution, the degree of correlation of the ith information in the input information vector X with the query vector q is shown given the query vector q.
5. The method for predicting fracture toughness of nickel-based alloy based on deep learning of claim 3, wherein after obtaining the second eigenvector, the method further comprises:
inputting the second feature vector into a BP neural network to obtain error information;
propagating the error information to the limited Boltzmann machine network of each layer by using a back propagation network, finishing fine adjustment of the depth confidence network based on the attention mechanism, and obtaining the fine-adjusted depth confidence network based on the attention mechanism;
and obtaining an optimized feature vector by using the fine-tuned depth confidence network based on the attention mechanism.
6. The deep learning-based nickel-based alloy fracture toughness prediction method as claimed in claim 1, wherein the trained support vector regression model comprises:
Figure FDA0003092982890000021
wherein C is a penalty factor, lεIs an ε -insensitive loss function, ε is a deviation threshold, W is a weight matrix, f (x)i) As output of the model, yiFor a true output, i is the count variable of the training samples, and m is the total number of training samples.
7. The deep learning-based nickel-base alloy fracture toughness prediction method as claimed in claim 1, wherein the kernel function of the trained support vector regression model is a radial basis kernel function.
8. The deep learning-based nickel-based alloy fracture toughness prediction method as claimed in claim 1, wherein the training method of the support vector regression model comprises:
and optimally selecting the optimal combination of the penalty factor and the radial basis function width of the support vector regression model by using a grid search algorithm in a K-fold cross validation mode to obtain the trained support vector regression model.
9. The deep learning-based nickel-base alloy fracture toughness prediction method as claimed in claim 8, wherein the optimization selection of the optimal combination of the penalty factor and the radial basis function width of the support vector regression model by using the grid search algorithm comprises:
listing all possible combinations of penalty factors and radial basis function widths of the support vector regression model, wherein all the possible combinations form a grid;
sequentially carrying out support vector regression modeling on possible combinations in the grid;
and selecting the possible combination with the highest modeling precision as the optimal combination of the penalty factor and the radial basis function width of the support vector regression model by using a k-fold cross verification method.
10. A deep learning based nickel-base alloy fracture toughness prediction system, the system comprising:
the input module is used for preprocessing the experimental data of the fracture toughness of the nickel-based alloy to obtain an experimental data set;
the deep learning module based on attention is used for learning and optimizing the features of the experimental data set by using a deep confidence network based on an attention mechanism to obtain an optimized feature vector;
the feature splicing module is used for splicing the optimized feature vector and the original feature vector of the experimental data set to obtain a predicted feature vector;
and the output module is used for predicting the fracture toughness of the nickel-based alloy through a trained support vector regression model according to the predicted characteristic vector.
CN202110601155.0A 2021-05-31 2021-05-31 Nickel-based alloy fracture toughness prediction method and system based on deep learning Active CN113343566B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110601155.0A CN113343566B (en) 2021-05-31 2021-05-31 Nickel-based alloy fracture toughness prediction method and system based on deep learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110601155.0A CN113343566B (en) 2021-05-31 2021-05-31 Nickel-based alloy fracture toughness prediction method and system based on deep learning

Publications (2)

Publication Number Publication Date
CN113343566A true CN113343566A (en) 2021-09-03
CN113343566B CN113343566B (en) 2023-09-01

Family

ID=77472834

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110601155.0A Active CN113343566B (en) 2021-05-31 2021-05-31 Nickel-based alloy fracture toughness prediction method and system based on deep learning

Country Status (1)

Country Link
CN (1) CN113343566B (en)

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106709640A (en) * 2016-12-15 2017-05-24 华南理工大学 Method for predicting ship energy consumption based on deep learning and support vector regression
CN109543352A (en) * 2018-12-29 2019-03-29 西南石油大学 Shale fracture toughness prediction technique, device and electronic equipment
CN110378035A (en) * 2019-07-19 2019-10-25 南京工业大学 It is a kind of that soft-measuring modeling method is hydrocracked based on deep learning
CN110852273A (en) * 2019-11-12 2020-02-28 重庆大学 Behavior identification method based on reinforcement learning attention mechanism
CN110929958A (en) * 2019-12-10 2020-03-27 西安邮电大学 Short-term traffic flow prediction method based on deep learning parameter optimization
CN111210633A (en) * 2020-02-09 2020-05-29 北京工业大学 Short-term traffic flow prediction method based on deep learning
CN111341386A (en) * 2020-02-17 2020-06-26 大连理工大学 Attention-introducing multi-scale CNN-BilSTM non-coding RNA interaction relation prediction method
WO2020144417A1 (en) * 2019-01-09 2020-07-16 Constellium Issoire Method for checking a damage tolerance property of a part made of an aluminium alloy
US20200257933A1 (en) * 2019-02-05 2020-08-13 Imagars Llc Machine Learning to Accelerate Alloy Design
US20210056678A1 (en) * 2019-08-22 2021-02-25 Saudi Arabian Oil Company System and method for cyber-physical inspection and monitoring of nonmetalic structures
CN112597921A (en) * 2020-12-28 2021-04-02 杭州电子科技大学 Human behavior recognition method based on attention mechanism GRU deep learning
CN112668804A (en) * 2021-01-11 2021-04-16 中国海洋大学 Method for predicting broken track of ground wave radar ship

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106709640A (en) * 2016-12-15 2017-05-24 华南理工大学 Method for predicting ship energy consumption based on deep learning and support vector regression
CN109543352A (en) * 2018-12-29 2019-03-29 西南石油大学 Shale fracture toughness prediction technique, device and electronic equipment
WO2020144417A1 (en) * 2019-01-09 2020-07-16 Constellium Issoire Method for checking a damage tolerance property of a part made of an aluminium alloy
US20200257933A1 (en) * 2019-02-05 2020-08-13 Imagars Llc Machine Learning to Accelerate Alloy Design
CN110378035A (en) * 2019-07-19 2019-10-25 南京工业大学 It is a kind of that soft-measuring modeling method is hydrocracked based on deep learning
US20210056678A1 (en) * 2019-08-22 2021-02-25 Saudi Arabian Oil Company System and method for cyber-physical inspection and monitoring of nonmetalic structures
CN110852273A (en) * 2019-11-12 2020-02-28 重庆大学 Behavior identification method based on reinforcement learning attention mechanism
CN110929958A (en) * 2019-12-10 2020-03-27 西安邮电大学 Short-term traffic flow prediction method based on deep learning parameter optimization
CN111210633A (en) * 2020-02-09 2020-05-29 北京工业大学 Short-term traffic flow prediction method based on deep learning
CN111341386A (en) * 2020-02-17 2020-06-26 大连理工大学 Attention-introducing multi-scale CNN-BilSTM non-coding RNA interaction relation prediction method
CN112597921A (en) * 2020-12-28 2021-04-02 杭州电子科技大学 Human behavior recognition method based on attention mechanism GRU deep learning
CN112668804A (en) * 2021-01-11 2021-04-16 中国海洋大学 Method for predicting broken track of ground wave radar ship

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
LING QIAO 等: "Alloy design and properties optimization of multi-component alloy based on solidification characteristics", 《MATERIALS SCIENCE AND ENGINEERING：A》, vol. 805, pages 1 - 10 *
YABIN XU 等: "A prediction Method of Fracture Toughness of Nickel-Based Superalloys", 《COMPUTER SYSTEMS SCIENCE & ENGINEERING》, vol. 42, no. 1, pages 121 - 132 *
YUE LIU 等: "Predicting creep rupture life of Ni-based single crystal superalloys using divide-and -conquer approach based machine learning", 《ACTA MATERIALIA》, vol. 195, pages 454 - 467, XP086220914, DOI: 10.1016/j.actamat.2020.05.001 *
汤宇豪;毛启容;高利剑;: "基于层次注意力机制的维度情感识别方法", 《计算机工程》, vol. 46, no. 06, pages 65 - 72 *
王太勇 等: "基于注意力机制BiLSTM的设备智能故障诊断方法", 《天津大学学报(自然科学与工程技术版)》, vol. 53, no. 06, pages 601 - 608 *
王太勇 等: "基于注意力机制BiLSTM的设备智能故障诊断方法", 天津大学学报(自然科学与工程技术版), vol. 53, no. 06, pages 601 - 608 *
陈远航: "滚动轴承剩余寿命预测算法研究及监测软件开发", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, no. 1, pages 029 - 355 *
韩朋 等: "基于AM-LSTM模型的超短期风电功率预测", 科学技术与工程, vol. 20, no. 21, pages 8594 - 8600 *

Also Published As

Publication number Publication date
CN113343566B (en) 2023-09-01

Similar Documents

Publication Publication Date Title
Mahalle et al. Neural network modeling for anisotropic mechanical properties and work hardening behavior of Inconel 718 alloy at elevated temperatures
KR101011546B1 (en) Prediction formula making device and prediction formula making method
Anijdan et al. Flow stress optimization for 304 stainless steel under cold and warm compression by artificial neural network and genetic algorithm
Chen et al. Efficient parameters identification of a modified GTN model of ductile fracture using machine learning
Hwang et al. Artificial intelligent analyzer for mechanical properties of rolled steel bar by using neural networks
Wang et al. High-throughput map design of creep life in low-alloy steels by integrating machine learning with a genetic algorithm
Gangi Setti et al. Artificial neural network approach for prediction of stress–strain curve of near β titanium alloy
CN110321658B (en) Method and device for predicting plate performance
CN106778846A (en) A kind of method for forecasting based on SVMs
Quan et al. Artificial neural network modeling to evaluate the dynamic flow stress of 7050 aluminum alloy
CN110472349A (en) A kind of hot-rolled steel performance prediction method based on EEMD and depth convolutional network
Wei et al. On the use of transfer modeling to design new steels with excellent rotating bending fatigue resistance even in the case of very small calibration datasets
Meng et al. Lightweight design of arcuately stiffened cylindrical shells based on smeared stiffener method and active learning strategy
Elbahy et al. Artificial neural network model for deflection analysis of superelastic shape memory alloy reinforced concrete beams
He et al. Multiaxial fatigue life prediction using physics-informed neural networks with sensitive features
Quan et al. Modelling the hot flow behaviors of AZ80 alloy by BP-ANN and the applications in accuracy improvement of computations
Mohanty et al. An adaptive neuro fuzzy inference system model for studying free in plane and out of plane vibration behavior of curved beams
CN113343566B (en) Nickel-based alloy fracture toughness prediction method and system based on deep learning
Mahdavi Jafari et al. Modeling and optimization of roll-bonding parameters for bond strength of Ti/Cu/Ti clad composites by artificial neural networks and genetic algorithm
Sun et al. Prediction of tensile property of hydrogenated Ti600 titanium alloy using artificial neural network
Xu et al. Prediction of static globularization of Ti-17 alloy with starting lamellar microstructure during heat treatment
Kessler et al. Incorporating neural network material models within finite element analysis for rheological behavior prediction
Ghiabakloo et al. Surrogate-based Pareto optimization of annealing parameters for severely deformed steel
Zhang et al. Rolling force prediction in heavy plate rolling based on uniform differential neural network
Zhen et al. Modeling constitutive relationship of Ti-555211 alloy by artificial neural network during high-temperature deformation

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant