CN116597923A

CN116597923A - Model generation method, material information determination method, device, equipment and medium

Info

Publication number: CN116597923A
Application number: CN202310575831.0A
Authority: CN
Inventors: 尹茸
Original assignee: Xiaomi Automobile Technology Co Ltd
Current assignee: Xiaomi Automobile Technology Co Ltd
Priority date: 2023-05-19
Filing date: 2023-05-19
Publication date: 2023-08-15

Abstract

The disclosure relates to a model generation method, a material information determination method, a device, equipment and a medium. The method comprises the following steps: determining alloy characteristic information of each of a plurality of alloy components by using at least one characteristic calculation unit, wherein the alloy characteristic information of one alloy component comprises characteristic values corresponding to a plurality of characteristic items, each characteristic calculation unit is used for determining the characteristic value of at least one characteristic item, one characteristic item is one of the plurality of characteristic items belonging to a preset category, the preset category comprises a structural category and a performance category, the same characteristic item in the alloy characteristic information is corresponding to at least one characteristic value, and each characteristic value is from one characteristic calculation unit; determining respective preset characteristic information of each alloy component, wherein the preset characteristic information at least comprises performance information; and generating a characteristic determination model according to the alloy characteristic information and the preset characteristic information, wherein the characteristic determination model is used for determining the preset characteristic information according to the alloy characteristic information.

Description

Model generation method, material information determination method, device, equipment and medium

Technical Field

The disclosure relates to the field of materials, and in particular relates to a model generation method, a material information determination method, a device, equipment and a medium.

Background

In the development of new materials and processes, there are usually involved a plurality of steps and links, for example, material calculation and simulation using a high-performance computer, laboratory preparation, etc., wherein a plurality of teams are involved, and meanwhile, since the calculation software and tools used by each team may be different, a great deal of data sharing, communication, coordination, integration, such as comparison and verification of calculation results of different processes, are required, and thus, a great deal of time and manpower resources are required to be consumed.

Disclosure of Invention

In order to overcome the problems in the related art, the present disclosure provides a model generation method, a material information determination method, a device, equipment and a medium.

According to a first aspect of embodiments of the present disclosure, there is provided a model generation method, the method including:

determining alloy characteristic information of each of a plurality of alloy components by using at least one characteristic calculation unit, wherein the alloy characteristic information of one alloy component comprises characteristic values corresponding to a plurality of characteristic items, each characteristic calculation unit is used for determining the characteristic value of at least one characteristic item, one characteristic item is one of the plurality of characteristic items belonging to a preset category, the preset category comprises a structural category and a performance category, the same characteristic item in the alloy characteristic information is corresponding to at least one characteristic value, and each characteristic value is from one characteristic calculation unit;

Determining preset characteristic information corresponding to each alloy component, wherein the preset characteristic information at least comprises performance information;

and generating a characteristic determination model according to the alloy characteristic information and the preset characteristic information, wherein the characteristic determination model is used for determining the preset characteristic information according to the alloy characteristic information.

Optionally, the feature calculation unit generates by:

determining a characteristic calculation result corresponding to the alloy component by using simulation calculation software according to material information of the alloy component, wherein the material information at least comprises component information;

and training the model by taking the material information as the input of the model and taking the characteristic calculation result corresponding to the material information as the target output of the model to obtain a trained model, wherein the trained model is taken as the characteristic calculation unit.

Optionally, the generating a feature determination model according to the alloy feature information and the preset feature information includes:

determining at least one characteristic item from a plurality of characteristic items contained in the alloy characteristic information, and taking the at least one characteristic item as a target characteristic item;

and performing model training by taking a characteristic value corresponding to a target characteristic item of the alloy component as an input of a model and taking preset characteristic information corresponding to the alloy component as a target output of the model to obtain the characteristic determination model after training.

Optionally, the determining at least one feature item from the plurality of feature items contained in the alloy feature information as a target feature item includes:

according to a plurality of characteristic items contained in the alloy characteristic information, determining the correlation degree between each characteristic item and the preset characteristic information;

and taking the first N characteristic items with high correlation degree as the target characteristic items, wherein N is a positive integer.

model training is carried out by taking alloy characteristic information as the input of a model and taking preset characteristic information corresponding to the alloy characteristic information as the target output of the model, so as to obtain a trained initial model;

in the initial model, determining the importance degree of each characteristic item in the alloy characteristic information on the preset characteristic information;

and taking the first M feature items with high importance as the target feature items, wherein M is a positive integer.

Optionally, the training the model by taking a feature value corresponding to a target feature item of an alloy component as an input of the model and taking preset feature information corresponding to the alloy component as a target output of the model to obtain the feature determination model after training, including:

Determining at least two first models;

training each first model according to the characteristic value corresponding to the target characteristic item of the alloy component and the preset characteristic information corresponding to the alloy component to obtain a second model corresponding to each first model;

and combining the second models into an integrated model according to a preset integrated learning mode, and training the integrated model to take the integrated model after training as the characteristic determining model, wherein the output result of the characteristic determining model is generated based on the output result of each second model.

Optionally, the alloy characteristic information further comprises characteristic values corresponding to component characteristic items and/or characteristic values corresponding to process condition characteristic items.

According to a second aspect of embodiments of the present disclosure, there is provided a material information determination method, the method including:

determining characteristic information of a target alloy;

inputting the target alloy characteristic information into a pre-generated characteristic determination model to obtain an output result of the characteristic determination model, wherein the output result comprises preset characteristic information corresponding to the target alloy characteristic information, and the characteristic determination model is generated based on the model generation method according to the first aspect of the disclosure.

According to a third aspect of embodiments of the present disclosure, there is provided a model generating apparatus, the apparatus including:

a first determining module configured to determine alloy characteristic information of each of a plurality of alloy components using at least one characteristic calculating unit, the alloy characteristic information of one alloy component including characteristic values corresponding to a plurality of characteristic items, each of the characteristic calculating units being configured to determine a characteristic value of at least one characteristic item, wherein one characteristic item is one of a plurality of characteristic items belonging to a preset category, the preset category including a structural category and a performance category, and the same one of the characteristic items in the alloy characteristic information corresponding to at least one characteristic value, each of the characteristic values being from one of the characteristic calculating units;

a second determining module configured to determine preset characteristic information corresponding to each of the alloy components, wherein the preset characteristic information at least comprises performance information;

the first processing module is configured to generate a feature determination model according to the alloy feature information and the preset feature information, wherein the feature determination model is used for determining the preset feature information according to the alloy feature information.

Optionally, the feature calculation unit generates by:

A third determination module configured to determine, using simulation calculation software, a feature calculation result corresponding to an alloy component based on material information of the alloy component, the material information including at least component information;

and the second processing module is configured to perform model training by taking the material information as an input of a model and taking a characteristic calculation result corresponding to the material information as a target output of the model to obtain a trained model, and take the trained model as the characteristic calculation unit.

Optionally, the first processing module includes:

a first determining submodule configured to determine at least one feature item from a plurality of feature items contained in the alloy feature information as a target feature item;

the first processing sub-module is configured to perform model training by taking a characteristic value corresponding to a target characteristic item of an alloy component as an input of a model and taking preset characteristic information corresponding to the alloy component as a target output of the model, so as to obtain the characteristic determination model after training.

Optionally, the first determining sub-module includes:

the second determining submodule is configured to determine the relativity between each characteristic item and the preset characteristic information according to a plurality of characteristic items contained in the alloy characteristic information;

And a third determining sub-module configured to use the first N feature items with high correlation degree as the target feature items, wherein N is a positive integer.

Optionally, the first determining sub-module includes:

the second processing submodule is configured to perform model training by taking alloy characteristic information as input of a model and taking preset characteristic information corresponding to the alloy characteristic information as target output of the model to obtain a trained initial model;

a fourth determining submodule configured to determine, in the initial model, importance of each feature item in the alloy feature information to the preset feature information;

and a fifth determining submodule configured to take the first M feature items with high importance as the target feature items, wherein M is a positive integer.

Optionally, the first processing sub-module includes:

a sixth determination submodule configured to determine at least two first models;

the third processing sub-module is configured to train each first model according to the characteristic value corresponding to the target characteristic item of the alloy component and the preset characteristic information corresponding to the alloy component to obtain a second model corresponding to each first model;

And the fourth processing submodule is configured to combine the second models into an integrated model according to a preset integrated learning mode, train the integrated model and take the integrated model after training as the characteristic determining model, wherein the output result of the characteristic determining model is generated based on the output result of each second model.

According to a fourth aspect of embodiments of the present disclosure, there is provided a material information determination apparatus, the apparatus including:

a fourth determination module configured to determine target alloy characteristic information;

and a third processing module configured to input the target alloy characteristic information into a pre-generated characteristic determination model, and obtain an output result of the characteristic determination model, wherein the output result comprises preset characteristic information corresponding to the target alloy characteristic information, and the characteristic determination model is generated based on the model generation method in the first aspect of the disclosure.

According to a fifth aspect of embodiments of the present disclosure, there is provided an electronic device, comprising:

A processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to perform the steps of the method of the first or second aspect of the present disclosure.

According to a sixth aspect of embodiments of the present disclosure, there is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the steps of the method provided in the first or second aspect of the present disclosure.

The technical scheme provided by the embodiment of the disclosure can comprise the following beneficial effects:

according to the technical scheme, at least one characteristic calculation unit is utilized to determine the respective alloy characteristic information of the plurality of alloy components, the respective corresponding preset characteristic information of each alloy component is determined, and then a characteristic determination model is generated according to the alloy characteristic information and the preset characteristic information, wherein the preset characteristic information at least comprises performance information. Therefore, the alloy characteristic information of the alloy component is determined through the characteristic calculation unit to generate a characteristic determination model capable of determining preset characteristic information of the alloy component according to the alloy characteristic information of the material, so that the characteristic determination model has the capability of predicting the preset characteristic of the material based on the alloy characteristic information of the material. The alloy characteristic information of the alloy component comprises characteristic values corresponding to various characteristic items, one characteristic item is one of various characteristic items belonging to preset categories, and the preset categories comprise structural categories and performance categories, so that the alloy characteristic information can contain abundant and various information, and besides the structural categories, the alloy characteristic information also comprises performance categories, namely, the performance can also serve as the characteristic items in the alloy characteristic information to serve as input of model training, thereby being beneficial to improving the diversity of data of training models and improving the performance of the models; meanwhile, the same characteristic item in the alloy characteristic information can be correspondingly provided with at least one characteristic value, each characteristic value comes from one characteristic calculation unit, namely, one characteristic item in the alloy characteristic information can be correspondingly provided with the characteristic value calculated by each calculation unit, and the model has stronger inclusion, so that the data of the training model is further enriched. Therefore, the data for training the feature determination model is enriched, so that the trained model is suitable for more prediction scenes, the prediction of the preset features of the material, particularly the prediction of the material performance, can be rapidly and efficiently realized based on the alloy feature information, and the method has higher accuracy compared with the direct prediction from the components to the performance which is commonly used at present. Based on the method, in the research and development process of new materials and processes, the preset characteristics of the inclusion performance of the new materials can be rapidly and accurately predicted through the characteristic determination model, so that the time and the occupation of human resources are greatly reduced.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure.

FIG. 1 is a flow chart illustrating a method of model generation according to an exemplary embodiment.

Fig. 2 is a flowchart illustrating a material information determination method according to an exemplary embodiment.

Fig. 3 is a block diagram of a model generating apparatus according to an exemplary embodiment.

Fig. 4 is a block diagram illustrating a material information determination apparatus according to an exemplary embodiment.

Fig. 5 is a block diagram illustrating an apparatus for performing a model generating method or for performing a material information determining method according to an exemplary embodiment.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.

It should be noted that, all actions for acquiring signals, information or data in the present disclosure are performed under the condition of conforming to the corresponding data protection rule policy of the country of the location and obtaining the authorization given by the owner of the corresponding device.

FIG. 1 is a flow chart illustrating a method of model generation according to an exemplary embodiment. As shown in fig. 1, the model generation method may include steps 11 to 13.

In step 11, alloy characteristic information of each of the plurality of alloy components is determined by at least one characteristic calculation unit.

Alloy characteristic information of one alloy composition includes characteristic values corresponding to a plurality of characteristic items. Wherein, a feature item may be one of a plurality of feature items belonging to a preset category, which may include, but is not limited to: structural class and performance class.

Alternatively, feature items belonging to a structural category may include, but are not limited to, the following:

band structure, electron density, charge density, crystal structure, scattering spectrum, crystal defects, phase diagram, component distribution, defect formation.

Alternatively, feature items belonging to a performance category may include, but are not limited to, the following:

dynamic behavior, thermodynamic properties, mechanical properties, surface properties, stress distribution, deformation fields, heat transfer, phase equilibrium, flow behavior, plastic deformation, phase transformation during casting.

The alloy characteristic information contains characteristic values of a plurality of characteristic items, and each characteristic item is taken from one of a plurality of items of a preset category, for example, all of the characteristic items mentioned above may be included. Therefore, the diversity and the richness of the alloy characteristic information can be ensured.

And, the same characteristic item in the alloy characteristic information can be corresponding to at least one characteristic value, and each characteristic value is from a characteristic calculation unit. That is, the alloy characteristic information may include a plurality of characteristic values corresponding to the same characteristic item, and each characteristic value is calculated based on a different characteristic calculation unit. For example, if the characteristic term of the energy band structure is calculated by the characteristic calculating unit K1, the energy band structure L2 of the alloy component is calculated by the characteristic calculating unit K2, and the energy band structure L3 of the alloy component is calculated by the characteristic calculating unit K3, the characteristic value of the characteristic term of the energy band structure is included in the alloy characteristic information, and the characteristic value of the characteristic term corresponding to the energy band structure includes L1, L2, and L3. Thus, a plurality of feature values which are similar to contradiction and correspond to the same feature item are allowed to be allowed, the content of the alloy feature information can be further enriched, and the diversity of the alloy feature information is improved.

In addition, the alloy characteristic information may further include characteristic values corresponding to constituent characteristic items and/or characteristic values corresponding to process condition characteristic items.

Characteristic values corresponding to characteristic items of the composition, that is, composition information of the alloy composition. The composition information corresponding to the alloy composition may include the composition and composition ratio of each raw material (e.g., the content of each metal in the total weight of the alloy raw material) in the alloy preparation experiment.

Characteristic values corresponding to characteristic items of process conditions, namely process condition information of alloy components. The process condition information of the alloy composition may include process conditions such as pressing pressure, temperature, pressing time, sintering temperature, sintering time, etc., which are involved in the alloy preparation conditions.

The characteristic value corresponding to the characteristic item in the alloy characteristic information is determined by the characteristic calculation unit. Wherein each feature calculation unit is used for determining a feature value of at least one feature item.

In one possible implementation, the feature calculation unit may be generated by:

according to the material information of the alloy components, determining a characteristic calculation result corresponding to the alloy components by using simulation calculation software;

the training is performed by taking the material information as the input of the model and taking the characteristic calculation result corresponding to the material information as the target output of the model, so that a trained model is obtained, and the trained model is taken as the characteristic calculation unit.

Wherein the material information includes at least component information. The material information of the alloy composition may include structural characterization parameters or the like of the alloy composition, which are known in the art to be available based on existing test means or the like, or parameters or the like required for ICME (Integrated Computational Materials Engineering, integrated computing materials engineering) calculations based on existing calculation software or calculation programs. Including but not limited to: crystal parameters (e.g., lattice constants, unit cell parameters, atomic radii, packing levels, coordination numbers, etc.), alloy molecular models (e.g., atomic coordinates, etc.), dislocation models (e.g., movement speed of dislocations, density of dislocations, etc.), phase field models (e.g., different orientations of grains, grain size distribution, grain morphology), alloy boundary conditions (e.g., alloy material geometry, stress intensity factors).

ICME integrates a tool for calculating material science into a whole and a material development process of a system, adopts calculation means (such as FT-Star casting software, proCAST simulation software and the like) to obtain material information, develops manufacturing process simulation, predicts the organization structure and the product performance of parts, realizes the calculation optimization of material components, manufacturing processes and components before actual manufacturing, and realizes the efficient development, manufacturing and use of advanced materials. By combining a multi-scale simulation and experimental method and integrating the system with a database, the development of the alloy can be gradually changed from an experience guidance experiment to a material research new mode of theoretical prediction and experimental verification.

The simulation calculation software may include any software involved in ICME (Integrated Computational Materials Engineering, integrated computing materials engineering) methods known in the art. For example, first principles computing software (e.g., VASP, quantum ESPRESSO, etc.), molecular dynamics simulation software (e.g., LAMMPS, GROMACS, etc.), crystallography software (e.g., X-Seed, VESTA, etc.), high throughput computing platform (e.g., materials Project, AFLOW, etc.), finite element analysis software (e.g., ANSYS, abaqus, etc.), data analysis and machine learning software (e.g., python, MATLAB, etc.), thermodynamic computing software (e.g., thermo-Calc, pandat, etc.), hydrodynamic simulation software (e.g., flow3D, magma, etc.).

Any structure or performance or behavior prediction of an alloy calculated by simulation calculation software known in the art may be used in the present disclosure as "alloy characteristic information" in the present disclosure, including but not limited to:

alloy characteristic information obtained by calculating according to information such as atomic coordinates, unit cell parameters, electron density and the like of the alloy by adopting first sexual principle calculation software: band structure (e.g., direct band gap, indirect band gap, electron effective mass, hole effective mass, total allowed bands, band width, total state density, etc.), electron density, charge density, crystal structure (e.g., crystal phase index, crystal plane index, interplanar spacing, crystal defects); first principles computing software including, but not limited to, VASP, quantum ESPRESSO, etc.;

And (3) calculating the obtained alloy characteristic information according to the atomic structure, the kinetic parameters and the like of the alloy by adopting molecular dynamics simulation software: dynamic behavior (e.g., alloy strain rate, yield limit, instantaneous stress, strain rate sensitivity limit, plastic deformation, shear band, dynamic fracture, etc.), thermodynamic properties (e.g., potential energy, specific surface energy, excess mixing entropy, excess mixing free energy, activity of each component, etc.), mechanical properties (e.g., hardness, brittleness, toughness, tensile strength, yield strength, ductility, stiffness, etc.), surface properties (e.g., high temperature oxidation properties, viscosity, surface tension, etc.); molecular dynamics simulation software includes, but is not limited to LAMMPS, GROMACS, etc.;

and (3) calculating alloy characteristic information according to the crystal structure parameters and the unit cell parameters of the alloy by adopting crystallographic software: crystal structure, scattering spectrum, crystal defects (e.g., grain boundaries, phase boundaries, dislocations, stacking structures, etc.), crystallographic software including, but not limited to, X-Seed, VESTA, etc.;

the property and behavior of the material are calculated by adopting a high-flux calculation platform and are used as alloy characteristic information, such as lattice information (and atomic coordinates) of an alloy, a band gap, a band structure of a crystal and the like; high-throughput computing platforms include, but are not limited to Materials Project, AFLOW, etc.;

Adopting finite element analysis software to calculate stress distribution, deformation field and heat transmission of the alloy as alloy characteristic information according to the confidence of the geometric structure, material parameters, boundary conditions and the like of the material; finite element analysis software includes, but is not limited to ANSYS, abaqus, and the like;

calculating to obtain the property and behavior prediction of the material as alloy characteristic information according to experimental data and calculation data of the material by adopting data analysis and machine learning software; data analysis and machine learning software including, but not limited to Python, MATLAB and the like;

calculating according to chemical components, temperature and pressure of the alloy by using thermodynamic calculation software to obtain phase diagrams, thermodynamic properties, phase balance and the like as alloy characteristic information; thermodynamic calculation software includes, but is not limited to Thermo-Calc, pandat;

calculating the flow behavior, heat transmission, plastic deformation, phase change in the casting process, component distribution, defect formation and the like of the alloy by adopting fluid mechanics simulation software according to the geometric structure, material parameters, boundary conditions and flow conditions of the alloy to obtain alloy characteristic information; hydrodynamic simulation software includes, but is not limited to, flow3D, magma, and the like.

In the present disclosure, the corresponding characteristic values may also be calculated according to the material information of the alloy components by calculation methods well known in the art. The calculation method and the calculation process can be performed through the model which is completed through training, but the method and the formula which are applied during the calculation of the model are well known in the art, and in the calculation process, the existing software can be introduced and the related calculation can be performed, and the software calculation process can also see a link in the calculation thought.

After the feature calculation result of the alloy component is obtained through the simulation calculation software, the model can be trained according to the material information of the alloy component and the feature calculation result so as to generate the feature calculation unit.

For example, a model constructed based on a neural network may be selected for the training. The training of the model aims to enable the feature output by the model to be infinitely close to the feature calculation result calculated by the simulation calculation software. Therefore, it is possible to calculate a loss value of the model based on the feature calculation result actually determined by the simulation calculation software and the result of the model output for the material information that has been input, and to adjust the internal parameters of the current model using the loss value. And then the adjusted model is used for the next training, and the model is circularly reciprocated until the condition of stopping training is met, so that the model after training is obtained, and the characteristic calculation unit is generated.

The feature calculation means generated in the above manner has the capability of generating the feature value of the alloy component from the material information of the alloy component, so that, for a certain alloy, after knowing the material information thereof, the feature value of the alloy corresponding to at least one feature item can be determined by the feature calculation means.

It should be noted that, the simulation calculation software may have the problems of low calculation efficiency and slow calculation, and by using the above-mentioned model training method, the actual feature calculation result of the simulation calculation software is used as the data set of the training model, so that the trained model (i.e. the feature calculation unit) can accurately determine the features of a certain or several feature items of the alloy, and at the same time, the efficiency of feature calculation can be ensured, for example, the single calculation can be reduced to millisecond level.

Therefore, for the calculation process that takes longer time to calculate the characteristic value of one or more characteristic items of the alloy in the actual operation process, the corresponding characteristic calculation unit can be generated in the mode.

By the method, the model is trained in a machine learning mode to generate the characteristic calculation unit, so that the characteristic calculation unit has both calculation accuracy and calculation speed, and the characteristic value of at least one characteristic item of the alloy can be quickly obtained.

In a further possible embodiment, the feature calculation unit may calculate the feature value of the at least one feature item by means of a predetermined calculation software or a predetermined calculation. The preset calculation mode may include, but is not limited to, formula calculation, empirical calculation, etc.

Alternatively, for a calculation process that takes a short time in the actual operation process, since the feature calculation result can be obtained quickly, it can be directly applied to the feature calculation unit. Or, for some calculation processes which can be specified through experience, the corresponding relation between input data and output data can be generated in advance and applied to the feature calculation unit, so that the calculation unit can quickly position the feature value of the corresponding feature item according to the corresponding relation based on the input material information.

Thus, in the at least one feature computing unit, each feature computing unit may be obtained by using any of the feature computing units provided in the present disclosure, for example, some feature computing units may be generated by using model training, and other feature computing units may be generated by using preset computing software or a preset computing method. The generation mode of the feature calculation unit can be flexibly selected and set according to actual requirements, and the method is not limited in this disclosure.

In step 12, the respective predetermined characteristic information for each alloy composition is determined.

Wherein the preset characteristic information at least comprises performance information.

In one possible embodiment, this can be achieved by means of at least one pre-set calculation unit which is generated beforehand, wherein each pre-set calculation unit is used to determine at least one pre-set characteristic of the alloy composition.

The properties of the alloy composition are data from conventional performance testing in the art, including physical and chemical properties of the alloy, and may include, but are not limited to: alloy physical properties, alloy chemical properties, and the like. Among other things, the alloy physical properties may include, but are not limited to: hardness density, strength, modulus, toughness, electrical conductivity, thermal conductivity, etc.; alloy chemistry may include, but is not limited to: corrosion resistance, aging resistance, and the like. Corresponding alloy performance data can be selected according to actual production requirements to construct a model.

Each preset calculating unit can determine at least one preset characteristic of the alloy component, so that according to at least one preset calculating unit generated in advance, a plurality of preset characteristics of any alloy component can be determined to be used as preset characteristic information corresponding to the alloy component.

The preset calculating unit may be a previously generated calculating unit that can rapidly calculate the preset characteristics of the alloy composition, and thus, the preset calculating unit can calculate the preset characteristics of the alloy composition by any available calculation means.

In one possible implementation, the preset calculating unit may be generated by:

Determining a preset characteristic calculation result corresponding to the alloy components by using simulation calculation software according to preset information of the alloy components;

and training the model by taking the preset information as the input of the model and taking the preset characteristic calculation result corresponding to the preset information as the target output of the model to obtain a trained model, wherein the trained model is taken as a calculation unit.

The preset information may be information related to the alloy including, but not limited to: composition information, process condition information, alloy characteristic information (or characteristic value of at least one characteristic item therein).

In the present disclosure, the corresponding alloy preset features may also be calculated by calculation methods well known in the art. The calculation method and the calculation process can be performed through the model which is completed through training, but the method and the formula which are applied during the calculation of the model are well known in the art, and in the calculation process, the existing software can be introduced and the related calculation can be performed, and the software calculation process can also see a link in the calculation thought.

The following are examples of several computing ideas, which may include, but are not limited to, the following in practical applications.

The first sexual principle calculation thought is as follows:

1. determining the lattice structure of an alloy material, wherein lattice structure data can be derived from an inorganic crystallography database, and input files required by density functional can be generated from the crystal structure, including automatic conversion of the crystal structure files, setting parameters such as k point, truncation energy, convergence accuracy, pseudo potential and the like;

2. calculating the electronic structure and the energy band structure of the alloy material by using a density functional theory; the density functional theory can calculate and obtain a plurality of physical properties including electronic energy bands, state density, binding energy, energy gaps, elastic constants and the like, and can calculate and obtain related data according to actual needs;

3. The thermodynamic property, elastic property, plastic property, fracture property and other mechanical properties of the alloy material are calculated by utilizing the energy band structure, and the related calculation can be performed by software.

The calculation thought of molecular dynamics simulation:

1. designing a molecular model of the alloy material, wherein the molecular model of the alloy material can be constructed by existing software, for example, according to parameters such as lattice constants, the number of atoms in a unit cell, the positions of the atoms and the like;

2. calculating the structure and the dynamics properties of the alloy material by a molecular dynamics simulation (MD simulation) method; the most direct research result of MD simulation is the structural characteristics of a molecular system, the MD simulation method can also research various thermodynamic properties of the molecular system, including kinetic energy, potential energy, enthalpy, gibbs free energy, hot melting and the like of the system, and the relation among density, pressure, volume, temperature and the like related to a state equation of the system can also be obtained through MD simulation; according to the energy and free energy of the system, the phase change and phase balance properties of the system can be directly or indirectly researched; in addition, the properties such as the velocity autocorrelation coefficient, the velocity correlation function, the mean square displacement and the like of the molecular system can be researched by utilizing MD simulation, and various migration properties such as the self-diffusion coefficient, the interdiffusion coefficient, the viscosity coefficient and the like of the system can be calculated; the properties related to chemical reactions such as chemical bond fracture and generation can be obtained by utilizing reactive molecular force field MD simulation or AIMD simulation;

3. And calculating thermodynamic properties, elastic properties, plastic properties, fracture properties and other mechanical properties of the alloy material by using molecular dynamics simulation results, and performing related calculation by using software.

Calculating thinking of dislocation mechanics simulation:

1. designing a lattice structure and a dislocation model of the alloy material;

2. calculating mechanical properties of the alloy material through dislocation mechanical simulation, including but not limited to dislocation energy, dislocation movement path, dislocation stacking energy and the like; for example, simulations were performed using creep-tie alloys based on microscopic dislocation motions for tensile, creep, cycling, creep fatigue interactions and anisotropy;

3. and calculating mechanical properties such as elastic property, plastic property, fracture property and the like of the alloy material by using dislocation mechanics simulation results, and performing related calculation by using software.

Calculating thought of phase field simulation:

1. designing a phase field model of the alloy material, wherein the phase field model can be obtained by constructing conventional software; the field model is generally divided into two major types, namely a continuous phase field and a microscopic phase field (discrete model), wherein the microscopic phase field model uses the probability of atoms occupying lattice positions as a field variable, and the simulation field is mainly focused on solid phase transformation, aging precipitation, martensitic transformation and the like; the field variable of the continuous phase field model is also called a phase field, and the function of the continuous phase field model is to avoid the difficulty brought by the tracking interface;

2. Phase transition behavior and phase boundary characteristics of the alloy material are calculated through phase field simulation; the diffusion interface is adopted in the phase field method, so that the difficulty of tracking the interface by the traditional sharp interface is avoided, and various complex microstructures can be subjected to two-dimensional and three-dimensional simulation; the phase field method may describe the microstructure evolution of the unbalanced process; in addition, the phase field model can be coupled with different external field equations to realize the combination of a macroscopic scale and a microscopic scale to simulate the microstructure evolution under the actions of a temperature field, a flow field, a magnetic field and the like, so that the influence of factors such as temperature gradient, flow field speed, supercooling degree, anisotropy and different preferred orientations on the microscopic morphology can be studied;

3. and calculating thermodynamic properties, elastic properties, plastic properties, fracture properties and other mechanical properties of the alloy material by using the phase field simulation result, and performing related calculation by using software.

Alternatively, simple machine learning may be performed, for example, to collect a large amount of data (e.g., composition, performance, etc.) to train the model.

After the preset feature calculation result is obtained through the simulation calculation software, the model can be trained according to the preset information of the alloy components and the preset feature calculation result so as to generate a preset calculation unit.

For example, a model constructed based on a neural network may be selected for the training. The training of the model aims at enabling the preset features output by the model to be infinitely close to the preset feature calculation results calculated by the simulation calculation software. Therefore, it is possible to calculate a loss value of the model based on the preset feature calculation result actually determined by the simulation calculation software and the result of the model output for the preset information of the alloy that has been input, and to adjust the internal parameters of the current model using the loss value. And then the adjusted model is used for the next training, and the model is circularly reciprocated until the condition of stopping training is met, so that the model after training is obtained, and the preset calculating unit is generated.

The preset calculating unit generated in the mode has the capability of generating the preset characteristic information of the alloy component according to the preset information of the alloy component, so that the preset characteristic information corresponding to the alloy can be generated after knowing the preset information of the alloy for a certain alloy.

It should be noted that, the simulation calculation software may have the problems of low calculation efficiency and slow calculation, and by using the foregoing model training manner, the actual calculation result of the preset feature of the simulation calculation software is used as the data set of the training model, so that the trained model (i.e., the preset calculation unit) can accurately determine the preset feature of the alloy, and meanwhile, the calculation efficiency of the preset feature can be ensured, for example, the single calculation can be reduced to millisecond level.

Therefore, for the calculation process that takes longer time to calculate the preset characteristic information of the alloy in the actual operation process, the corresponding preset calculation unit can be generated in the mode.

Through the mode, the model is trained in a machine learning mode to generate the preset calculating unit, so that the calculating unit has both calculating accuracy and calculating speed, and the preset characteristic information of the alloy can be obtained rapidly.

In another possible embodiment, the computing unit may be configured to instruct a preset computing software or a preset computing mode. The preset calculation mode may include, but is not limited to, formula calculation, empirical calculation, etc.

Alternatively, for a calculation process that takes a relatively short time in an actual operation process, since the preset feature calculation result can be obtained quickly, it can be directly applied to the preset calculation unit. Or, for some calculation processes which can be specified through experience, the corresponding relation between the input data and the output data can be generated in advance and applied to a preset calculation unit, so that the preset calculation unit can quickly position corresponding preset characteristic information according to the corresponding relation based on the input information.

Thus, in the at least one preset computing unit, each preset computing unit may be obtained by using any of the preset computing units provided in the present disclosure, for example, some preset computing units may be generated by using a model training manner, and other preset computing units may be generated by using preset computing software or a preset computing manner. The generation mode of the preset calculation unit can be flexibly selected and set according to actual requirements, and the method is not limited in this disclosure.

In step 13, a feature determination model is generated based on the alloy feature information and the preset feature information.

The characteristic determining model is used for determining preset characteristic information according to the alloy characteristic information.

Alternatively, step 13 may comprise the steps of:

and performing model training by taking a characteristic value corresponding to a target characteristic item of the alloy component as an input of the model and taking preset characteristic information corresponding to the alloy component as a target output mode of the model to obtain a characteristic determination model after training.

Because the alloy characteristic information contains various characteristic values of various characteristic items, the correlation degree of each characteristic item and the preset characteristic prediction is unknown, some correlation degrees are high, some correlation degrees are low, and the characteristic with low correlation degree has little contribution to the prediction of the preset characteristic and even noise is introduced, the characteristic items in the alloy characteristic information can be selected, and the characteristic items which are more favorable for improving the accuracy of the preset characteristic can be screened out. The method of feature selection may employ filtering, packaging, embedding, etc., and the disclosure may employ any feature selection manner to determine the target feature item.

In one possible embodiment, at least one feature item is determined from a plurality of feature items contained in the alloy feature information, and the method may include the following steps as a target feature item:

according to a plurality of characteristic items contained in the alloy characteristic information, determining the correlation degree between each characteristic item and preset characteristic information;

and taking the first N characteristic items with high correlation degree as target characteristic items.

Wherein N is a positive integer.

This approach corresponds to a filtering method of feature selection, where the determination of the degree of correlation may use a statistical method or a correlation method. The method has the advantages that the degree of correlation is high, the degree of correlation between the characteristic item and the preset characteristic information is higher, the accurate prediction of the preset characteristic is facilitated, on the contrary, the degree of correlation is low, the degree of correlation between the characteristic item and the preset characteristic prediction is not high, and positive influence on the accurate prediction of the preset characteristic cannot be generated.

Therefore, after the relevance between each feature item and the preset feature information is determined, the top N feature items are determined as target feature items for training the feature determination model, and the feature items can be ranked in the order of the relevance from high to low.

In another possible embodiment, at least one feature item is determined from a plurality of feature items included in the alloy feature information, and the method may include the steps of:

Model training is carried out by taking alloy characteristic information as the input of a model and taking preset characteristic information corresponding to the alloy characteristic information as the target output of the model, so as to obtain an initial model after training;

in the initial model, determining the importance degree of each characteristic item in the alloy characteristic information on preset characteristic information;

the first M feature items with high importance are taken as target feature items.

Wherein M is a positive integer.

This approach corresponds to the feature importance training approach for evaluating the extent of influence of each feature on the target variable in the trained model, which can be reflected by the importance. Based on the above, the higher the importance degree is, the larger the influence of the feature item on the preset feature is, the more accurate prediction of the preset feature is facilitated, and on the contrary, the lower the importance degree is, the smaller the influence of the feature item on the preset feature is, and the prediction contribution of the feature item on the preset feature is not large. For example, determining the importance of each feature item to the preset feature information in the initial model may be implemented in a random forest, a gradient lifting tree, a linear regression, or the like.

Therefore, after determining the importance of each feature item to the preset feature information, the first M feature items may be determined as target feature items for training the feature determination model by sorting the feature items in order of importance from high to low.

As described above, the alloy characteristic information may further include a characteristic value corresponding to the component characteristic term and/or a characteristic value corresponding to the process condition characteristic term, and thus the target characteristic term may also include the component characteristic term or the process condition characteristic term, that is, the component information, the process condition information may also be input as the characteristic determination model.

After the target feature item is determined, training of the model can be performed, namely, training of the model is performed by taking a feature value corresponding to the target feature item of the alloy component as an input of the model and taking preset feature information corresponding to the alloy component as a target output of the model, so that a feature determination model after training is completed is obtained.

In one possible implementation, the training may be performed using a model constructed based on a neural network. The training of the model aims at enabling the preset features output by the model to be infinitely close to the preset feature information of the alloy. Therefore, the loss value of the model can be calculated based on the preset feature information determined in step 12 and the result output by the model for the feature value corresponding to the target feature item of the inputted alloy component, and the internal parameter of the current model can be adjusted by using the loss value. And then the adjusted model is used for the next training, and the model is repeatedly used until the condition of stopping training is met, so that the characteristic determination model after training is completed can be obtained.

The characteristic determination model generated in the mode has the capability of generating the preset characteristic information of the material according to the characteristic value of the characteristic item of the material, so that the preset characteristic information corresponding to the alloy can be generated after the alloy characteristic information of the alloy is known.

In another possible embodiment, the feature determination model described above may be trained by means of ensemble learning. Thus, the training is performed by taking the feature value corresponding to the target feature item of the alloy component as the input of the model and taking the preset feature information corresponding to the alloy component as the target output of the model, so as to obtain the feature determination model after training, and the method may include the following steps:

determining at least two first models;

and combining the second model into an integrated model according to a preset integrated learning mode, and training the integrated model to take the integrated model after training as a characteristic determination model.

Wherein the output results of the feature determination models are generated based on the output results of the respective second models.

The first models, i.e. the base model in ensemble learning, may be any type of model and the type of each first model need not be identical. For example, the first model may be any one of decision tree, support vector machine, and neural network.

After the first models are determined, each first model can be trained based on the characteristic value corresponding to the target characteristic item and the preset characteristic information, so that the model has the capability of accurately predicting the preset characteristic information of the alloy according to the characteristic value corresponding to the target characteristic item in training. During training, each first model can be used as an independent predictor for training based on the characteristic value corresponding to the target characteristic item and the corresponding preset characteristic information in the collected alloy characteristic information, so that a trained model, namely a second model, is obtained.

Based on the method, an integration method is used, a plurality of second models are combined into one integration model according to a preset integration learning mode, and meanwhile, the prediction result of each second model is combined into one integration prediction. In forming the integrated model, the superparameter of the integrated model, such as the number of second models, the weight of each second model, etc., may be set by means of cross-validation, etc. And training the integrated model based on the feature value corresponding to the target feature item and the preset feature information, and adjusting the super-parameters of the integrated model in the training process until the training is completed to obtain the feature determination model. The preset ensemble learning manner may include, but is not limited to, voting, averaging, weighted averaging, stacking, and the like.

By means of the method, the model is trained by means of the integrated learning mode to generate the feature determination model, and prediction accuracy of the feature determination model can be further improved.

The method and the device have the advantages that structural information of the alloy is not limited to be used as input, calculated performance information and the like can be used as input, the diversity of the input information is greatly enriched, various information is connected with the finally output alloy prediction preset characteristics, and the accuracy of the final output is improved. In addition, the method can obtain the same kind of alloy characteristic information by adopting different calculation software or calculation methods, for example, the elastic modulus 1 is obtained by calculating the software 1, the elastic modulus 2 is obtained by calculating the software 2, and the quantitative selection is the elastic modulus property of the alloy, but the elastic modulus 1 and the elastic modulus 2 of the calculation results output by the two kinds of software can be the same or different, even if the calculation results are different, the two different calculation results can be used as input information at the same time, so that the method provided by the method can allow contradictory data to be input at the same time, and has stronger inclusion.

Fig. 2 is a flowchart illustrating a material information determination method according to an exemplary embodiment. As shown in fig. 2, the material information determination method may include steps 21 and 22.

In step 21, target alloy characteristic information is determined.

In step 22, the target alloy characteristic information is input into a characteristic determination model which is generated in advance, and an output result of the characteristic determination model is obtained.

The target alloy characteristic information may include, among other things, characteristic values corresponding to target characteristic items (i.e., target characteristic items for generating a characteristic determination model to be used), for example, in a development stage of a new material, characteristic values corresponding to target characteristic items of a new material to be prepared may be taken as target alloy characteristic information. The target feature item is described in the foregoing, and is not described here.

The output result includes preset feature information corresponding to the target alloy feature information, the preset feature information including at least the performance feature. The feature determination model is generated based on the model generation method provided by any embodiment of the present disclosure.

Furthermore, based on the feature determination model provided by any embodiment of the present disclosure, the feature determination model has the capability of predicting the preset feature including the performance according to the alloy feature information, so that the target alloy feature information can be input into the feature determination model generated in advance, and an output result of the feature determination model is obtained, where the output result is used for characterizing the preset feature of the new material predicted by the feature determination model based on the target alloy feature information.

Based on the method, the preset characteristics of the inclusion performance of the material expected to be prepared can be rapidly predicted without actually preparing the material, so that the time for waiting for the preparation of the material is saved, and whether the new material can be put into production or not can be rapidly judged. For example, if the output result predicted by the feature determination model does not reach the desire for a certain preset feature, the preparation mode (such as component or process condition) needs to be adjusted in time; for another example, if the output result predicted by the feature determination model reaches the desire of the preset feature, the preparation of the new material can be directly performed by using the current preparation mode. Thus, the research and development of new materials can be guided.

Through the technical scheme, the characteristic determination model is utilized, and the preset characteristic information which at least contains the performance and corresponds to the target alloy characteristic information can be predicted through the target alloy characteristic information, so that whether the target alloy characteristic information meets the requirements or not can be rapidly determined according to the predicted preset characteristic result, the preparation scheme of the new material can be adjusted in time, and the research and development efficiency of the new material is greatly improved.

Fig. 3 is a block diagram of a model generating apparatus according to an exemplary embodiment. Referring to fig. 3, the model generating apparatus 30 includes:

A first determining module 31 configured to determine alloy characteristic information of each of a plurality of alloy components using at least one characteristic calculating unit, the alloy characteristic information of one alloy component including characteristic values corresponding to a plurality of characteristic items, each of the characteristic calculating units being configured to determine a characteristic value of at least one characteristic item, wherein one characteristic item is one of a plurality of characteristic items belonging to a preset category, the preset category including a structural category and a performance category, and the same one of the characteristic items in the alloy characteristic information corresponding to at least one characteristic value, each of the characteristic values being from one of the characteristic calculating units;

a second determining module 32 configured to determine preset characteristic information corresponding to each of the alloy compositions, the preset characteristic information including at least performance information;

the first processing module 33 is configured to generate a feature determination model according to the alloy feature information and the preset feature information, wherein the feature determination model is used for determining the preset feature information according to the alloy feature information.

Optionally, the feature calculation unit generates by:

Optionally, the first processing module 33 includes:

Optionally, the first determining sub-module includes:

Optionally, the first processing sub-module includes:

Fig. 4 is a block diagram illustrating a material information determination apparatus according to an exemplary embodiment. Referring to fig. 4, the material information determining apparatus 40 includes:

a fourth determination module 41 configured to determine target alloy characteristic information;

a third processing module 42, configured to input the target alloy feature information into a pre-generated feature determination model, and obtain an output result of the feature determination model, where the output result includes preset feature information corresponding to the target alloy feature information, and the feature determination model is generated based on the model generation method provided by any embodiment of the present disclosure.

The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.

The present disclosure also provides an electronic device, including:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to perform the steps of the model generation method or the material information determination method provided by the present disclosure.

The present disclosure also provides a computer-readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the steps of the model generation method or the material information determination method provided by the present disclosure.

The present disclosure also provides a chip comprising a processor and an interface; the processor is configured to read the instructions to perform the steps of the model generation method or the material information determination method provided by the present disclosure.

Fig. 5 is a block diagram illustrating an apparatus 800 for performing a model generation method or for performing a material information determination method, according to an example embodiment. For example, apparatus 800 may be a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, exercise device, personal digital assistant, or the like.

Referring to fig. 5, apparatus 800 may include one or more of the following components: a processing component 802, a memory 804, a power component 806, a multimedia component 808, an audio component 810, an input/output interface 812, a sensor component 814, and a communication component 816.

The processing component 802 generally controls overall operation of the apparatus 800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to perform all or part of the steps of the model generation method or the material information determination method described above. Further, the processing component 802 can include one or more modules that facilitate interactions between the processing component 802 and other components. For example, the processing component 802 can include a multimedia module to facilitate interaction between the multimedia component 808 and the processing component 802.

The memory 804 is configured to store various types of data to support operations at the apparatus 800. Examples of such data include instructions for any application or method operating on the device 800, contact data, phonebook data, messages, pictures, videos, and the like. The memory 804 may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.

The power supply component 806 provides power to the various components of the device 800. The power components 806 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for the device 800.

The multimedia component 808 includes a screen between the device 800 and the user that provides an output interface. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may sense not only the boundary of a touch or slide action, but also the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and/or a rear camera. The front camera and/or the rear camera may receive external multimedia data when the apparatus 800 is in an operational mode, such as a photographing mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a Microphone (MIC) configured to receive external audio signals when the device 800 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may be further stored in the memory 804 or transmitted via the communication component 816. In some embodiments, audio component 810 further includes a speaker for outputting audio signals.

Input/output interface 812 provides an interface between processing component 802 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: homepage button, volume button, start button, and lock button.

The sensor assembly 814 includes one or more sensors for providing status assessment of various aspects of the apparatus 800. For example, the sensor assembly 814 may detect an on/off state of the device 800, a relative positioning of the components, such as a display and keypad of the device 800, the sensor assembly 814 may also detect a change in position of the device 800 or a component of the device 800, the presence or absence of user contact with the device 800, an orientation or acceleration/deceleration of the device 800, and a change in temperature of the device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscopic sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 816 is configured to facilitate communication between the apparatus 800 and other devices, either in a wired or wireless manner. The device 800 may access a wireless network based on a communication standard, such as WiFi,2G or 3G, or a combination thereof. In one exemplary embodiment, the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication component 816 further includes a Near Field Communication (NFC) module to facilitate short range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, ultra Wideband (UWB) technology, bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the apparatus 800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic elements for performing the above-described model generation method or material information determination method.

In an exemplary embodiment, a non-transitory computer readable storage medium is also provided, such as memory 804 including instructions executable by processor 820 of apparatus 800 to perform the model generation method or the material information determination method described above. For example, the non-transitory computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

The apparatus may be a stand-alone electronic device or may be part of a stand-alone electronic device, for example, in one embodiment, the apparatus may be an integrated circuit (Integrated Circuit, IC) or a chip, where the integrated circuit may be an IC or may be a collection of ICs; the chip may include, but is not limited to, the following: GPU (Graphics Processing Unit, graphics processor), CPU (Central Processing Unit ), FPGA (Field Programmable Gate Array, programmable logic array), DSP (Digital Signal Processor ), ASIC (Application Specific Integrated Circuit, application specific integrated circuit), SOC (System on Chip, SOC, system on Chip or System on Chip), etc. The integrated circuit or chip may be configured to execute executable instructions (or code) to implement the model generation method or the material information determination method described above. The executable instructions may be stored on the integrated circuit or chip or may be retrieved from another device or apparatus, such as the integrated circuit or chip including a processor, memory, and interface for communicating with other devices. The executable instructions may be stored in the memory, which when executed by the processor implement the model generation method or the material information determination method described above; alternatively, the integrated circuit or chip may receive executable instructions through the interface and transmit the executable instructions to the processor for execution to implement the model generation method or the material information determination method described above.

In another exemplary embodiment, a computer program product is also provided, comprising a computer program executable by a programmable apparatus, the computer program having code portions for performing the above-described model generation method or material information determination method when executed by the programmable apparatus.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure. This disclosure is intended to cover any adaptations, uses, or adaptations of the disclosure following the general principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It is to be understood that the present disclosure is not limited to the precise arrangements and instrumentalities shown in the drawings, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method of generating a model, the method comprising:

2. The method according to claim 1, wherein the feature calculation unit generates by:

3. The method of claim 1, wherein generating a feature determination model from the alloy feature information and the preset feature information comprises:

4. A method according to claim 3, wherein said determining at least one feature item from a plurality of feature items included in said alloy feature information as a target feature item comprises:

5. A method according to claim 3, wherein said determining at least one feature item from a plurality of feature items included in said alloy feature information as a target feature item comprises:

6. A method according to claim 3, wherein the training is performed by taking a feature value corresponding to a target feature item of an alloy component as an input of a model, and taking preset feature information corresponding to the alloy component as a target output of the model, to obtain the feature determination model after training is completed, and the method comprises:

determining at least two first models;

7. The method according to any one of claims 1-6, wherein the alloy characteristic information further comprises characteristic values corresponding to constituent characteristic items and/or characteristic values corresponding to process condition characteristic items.

8. A method of determining material information, the method comprising:

determining characteristic information of a target alloy;

inputting the target alloy characteristic information into a pre-generated characteristic determination model to obtain an output result of the characteristic determination model, wherein the output result comprises preset characteristic information corresponding to the target alloy characteristic information, and the characteristic determination model is generated based on the model generation method of any one of claims 1-7.

9. A model generation apparatus, characterized in that the apparatus comprises:

10. A material information determination apparatus, characterized in that the apparatus comprises:

a third processing module configured to input the target alloy characteristic information into a pre-generated characteristic determination model, and obtain an output result of the characteristic determination model, wherein the output result comprises preset characteristic information corresponding to the target alloy characteristic information, and the characteristic determination model is generated based on the model generation method according to any one of claims 1-7.

11. An electronic device, comprising:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to perform the steps of the method of any of claims 1-8.

12. A computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the steps of the method of any of claims 1-8.