CN116665823A - Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium - Google Patents
Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116665823A CN116665823A CN202310686934.4A CN202310686934A CN116665823A CN 116665823 A CN116665823 A CN 116665823A CN 202310686934 A CN202310686934 A CN 202310686934A CN 116665823 A CN116665823 A CN 116665823A
- Authority
- CN
- China
- Prior art keywords
- crystal structure
- predicted
- initial
- energy
- prediction model
- 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.)
- Pending
Links
- 239000013078 crystal Substances 0.000 title claims abstract description 438
- 238000000034 method Methods 0.000 title claims abstract description 84
- 239000000463 material Substances 0.000 claims abstract description 153
- 238000012549 training Methods 0.000 claims abstract description 51
- 238000005457 optimization Methods 0.000 claims abstract description 37
- 238000003062 neural network model Methods 0.000 claims abstract description 35
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000000605 extraction Methods 0.000 claims description 31
- 238000004590 computer program Methods 0.000 claims description 17
- 238000005381 potential energy Methods 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 11
- 238000012163 sequencing technique Methods 0.000 claims description 8
- 238000013528 artificial neural network Methods 0.000 abstract description 10
- 238000003672 processing method Methods 0.000 abstract description 4
- 125000004429 atom Chemical group 0.000 description 43
- 230000006870 function Effects 0.000 description 10
- 238000004891 communication Methods 0.000 description 9
- 102100021164 Vasodilator-stimulated phosphoprotein Human genes 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 108010054220 vasodilator-stimulated phosphoprotein Proteins 0.000 description 5
- 238000003775 Density Functional Theory Methods 0.000 description 3
- 125000004122 cyclic group Chemical group 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical group [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 2
- FAPWRFPIFSIZLT-UHFFFAOYSA-M Sodium chloride Chemical compound [Na+].[Cl-] FAPWRFPIFSIZLT-UHFFFAOYSA-M 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 125000001309 chloro group Chemical group Cl* 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 125000004430 oxygen atom Chemical group O* 0.000 description 2
- 125000004436 sodium atom Chemical group 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 229910052729 chemical element Inorganic materials 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000008204 material by function Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 235000012239 silicon dioxide Nutrition 0.000 description 1
- 239000000377 silicon dioxide Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000011780 sodium chloride Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N20/00—Machine learning
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a crystal structure prediction method, a crystal structure prediction device, electronic equipment and a storage medium. The method comprises the following steps: acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted, further training an initial neural network model based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model, and performing crystal parameter optimization processing on the crystal structure prediction model to finally obtain target crystal structure parameters. According to the technical scheme, the neural network prediction model and the parameter optimization processing method are combined to predict the crystal structure, so that the global optimal solution of the crystal structure is obtained, and compared with the traditional crystal structure prediction method, the crystal structure prediction operation time is shortened, and therefore the efficiency of predicting the crystal structure is improved.
Description
Technical Field
The present invention relates to the field of material structures, and in particular, to a crystal structure prediction method, a crystal structure prediction device, an electronic device, and a storage medium.
Background
The material is a material basis for promoting the development of technology, and the crystal structure prediction is a fundamental task of the invention and discovery of novel functional materials.
In the existing scheme, a new crystal structure can be predicted by a density functional theory (Density functional theory, DFT), and the method has the problems of long calculation time and low efficiency of the predicted crystal structure.
Disclosure of Invention
The invention provides a crystal structure prediction method, a crystal structure prediction device, electronic equipment and a storage medium, so as to improve the efficiency of predicting a crystal structure.
According to an aspect of the present invention, there is provided a crystal structure prediction method, including:
acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted;
training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model;
and carrying out crystal parameter optimization treatment on the crystal structure prediction model to obtain target crystal structure parameters.
According to another aspect of the present invention, there is provided a crystal structure prediction apparatus, comprising:
The training data acquisition module is used for acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted;
the prediction model training module is used for training the initial neural network model based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model;
and the crystal structure parameter acquisition module is used for carrying out crystal parameter optimization processing on the crystal structure prediction model to obtain target crystal structure parameters.
According to another aspect of the present invention, there is provided an electronic apparatus including:
at least one processor;
and a memory communicatively coupled to the at least one processor;
wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of predicting crystal structure according to any one of the embodiments of the present invention.
According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to execute a method of predicting a crystal structure according to any one of the embodiments of the present invention.
According to the technical scheme, the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted are obtained, and then the initial neural network model is trained based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted, so that a crystal structure prediction model is obtained, and crystal parameter optimization processing is carried out on the crystal structure prediction model, so that target crystal structure parameters are obtained. According to the technical scheme, the neural network prediction model and the parameter optimization processing method are combined to predict the crystal structure, so that the global optimal solution of the crystal structure is obtained, and compared with the traditional crystal structure prediction method, the crystal structure prediction operation time is shortened, and therefore the efficiency of predicting the crystal structure is improved.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a method for predicting a crystal structure according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a method for predicting a crystal structure according to a second embodiment of the present invention;
FIG. 3 is a flowchart of a method for predicting a crystal structure according to a third embodiment of the present invention;
FIG. 4 is a flowchart of a method for predicting a crystal structure according to a fourth embodiment of the present invention;
FIG. 5 is a flowchart of a method for predicting a crystal structure according to a fifth embodiment of the present invention;
FIG. 6 is a flowchart of a method for predicting a crystal structure according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of a crystal structure prediction apparatus according to a sixth embodiment of the present invention;
fig. 8 is a schematic structural diagram of an electronic device implementing a crystal structure prediction method according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
Fig. 1 is a flowchart of a crystal structure prediction method according to an embodiment of the present invention, where the method may be performed by a crystal structure prediction device of a novel functional material, the crystal structure prediction device may be implemented in hardware and/or software, and the crystal structure prediction device may be configured in a terminal and/or a server. As shown in fig. 1, the method includes:
S110, acquiring an initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted.
In this embodiment, the material to be predicted refers to a material to be subjected to crystal structure prediction, and the material to be predicted may be composed of a plurality of chemical elements. By way of example, the material to be predicted may be a material composed of chlorine atoms and sodium atoms, or a material composed of oxygen atoms and silicon atoms, or the like. The initial crystal structure of the material to be predicted may include a variety of types of crystal structures, which types may include, but are not limited to, simple cubes, body centered cubes, face centered cubes, and the like. The energy label corresponding to the initial crystal structure refers to the total energy of the initial crystal structure, wherein the total energy of the initial crystal structure is composed of potential energy and kinetic energy of atoms in the crystal, and the total energy can be calculated by a computer program package (Vienna Ab-initio Simulation Package, VASP) for simulating an atomic scale material.
Specifically, an initial crystal structure of a material to be predicted and an energy tag corresponding to the initial crystal structure of the material to be predicted can be fetched from a preset storage path of the electronic device; the initial crystal structure of the material to be predicted and the energy tag corresponding to the initial crystal structure of the material to be predicted may also be obtained from other devices or cloud connected with the electronic device in a communication manner, which is not limited herein.
S120, training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model.
The initial neural network model may be an untrained neural network model or a pre-trained neural network model, and the network architecture may be any architecture, which is not limited herein. Specifically, feature extraction is performed on an initial crystal structure of a material to be predicted through a neural network model to obtain atomic features of the initial crystal structure, prediction is performed on the basis of the atomic features of the initial crystal structure to obtain predicted energy, and parameter adjustment is performed on the initial neural network model on the basis of the predicted energy and the loss of an energy label until a model training stop condition is met to obtain a trained crystal structure prediction model.
S130, performing crystal parameter optimization processing on the crystal structure prediction model to obtain target crystal structure parameters.
In this embodiment, the crystal parameter optimization process refers to a super parameter optimization process, in which a crystal structure prediction model is used as an optimization target of the super parameter optimization process, and the target crystal structure parameter is a global optimal solution of the crystal structure prediction model.
It should be noted that, by performing crystal parameter optimization processing on the crystal structure prediction model, a global optimal solution of the crystal structure prediction model can be obtained, thereby improving accuracy of predicting the crystal structure.
According to the technical scheme, the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted are obtained, and then the initial neural network model is trained based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted, so that a crystal structure prediction model is obtained, and crystal parameter optimization processing is carried out on the crystal structure prediction model, so that target crystal structure parameters are obtained. According to the technical scheme, the neural network prediction model and the parameter optimization processing method are combined to predict the crystal structure, so that the optimal solution of the crystal structure is obtained, and compared with the traditional crystal structure prediction method, the crystal structure prediction operation time is shortened, and therefore the efficiency of predicting the crystal structure is improved.
Example two
Fig. 2 is a flowchart of a crystal structure prediction method according to a second embodiment of the present invention, where the method according to the present embodiment may be combined with each of the alternatives in the crystal structure prediction method according to the foregoing embodiment. The crystal structure prediction method provided in this embodiment is further optimized. Optionally, the obtaining the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted includes: acquiring component information and proportioning information of a material to be predicted; generating a plurality of initial crystal structures of the material to be predicted based on the component information and the proportioning information of the material to be predicted; and determining energy of each initial crystal structure of the material to be predicted, and obtaining energy labels corresponding to each initial crystal structure of the material to be predicted.
As shown in fig. 2, the method includes:
s210, component information and proportioning information of the materials to be predicted are obtained.
The material to be predicted can be sodium chloride, the component information comprises chlorine atoms and sodium atoms, and the proportion information is one to one; the material to be predicted can be silicon dioxide, the component information comprises oxygen atoms and silicon atoms, and the proportion information is two-to-one.
S220, generating various initial crystal structures of the material to be predicted based on the composition information and the proportioning information of the material to be predicted.
The initial crystal structure may be a crystal structure randomly generated according to composition information and proportioning information of the material to be predicted.
For example, atoms may be placed at Wyckoff locations according to composition information and proportioning information of the material to be predicted, and the placing and merging processes may be repeated for pairs of atoms that are present in close proximity until the stoichiometry of the material to be predicted is reached, resulting in a plurality of initial crystal structures of the material to be predicted. Wherein the Wyckoff position is used to represent symmetry of the equivalent atoms in the unit cell.
S230, determining energy of each initial crystal structure of the material to be predicted, and obtaining energy labels corresponding to each initial crystal structure of the material to be predicted.
Specifically, total energy calculation can be performed on each initial crystal structure through VASP software, so that total energy corresponding to each initial crystal structure is obtained, and the total energy corresponding to each initial crystal structure is used as an energy label corresponding to each initial crystal structure.
S240, training an initial neural network model based on the initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model.
S250, performing crystal parameter optimization processing on the crystal structure prediction model to obtain target crystal structure parameters.
According to the technical scheme, the component information and the proportion information of the material to be predicted are obtained, and then various initial crystal structures of the material to be predicted are generated based on the component information and the proportion information of the material to be predicted, and then the total energy of all the initial crystal structures of the material to be predicted is determined, so that energy labels corresponding to all the initial crystal structures of the material to be predicted are obtained, training data of a model are generated, and a data base is provided for model training.
Example III
Fig. 3 is a flowchart of a crystal structure prediction method according to a third embodiment of the present invention, where the method according to the present embodiment may be combined with each of the alternatives in the crystal structure prediction method according to the foregoing embodiment. The crystal structure prediction method provided in this embodiment is further optimized. Optionally, training the initial neural network model based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model, including: extracting features of the initial crystal structure of the material to be predicted to obtain atomic features corresponding to the initial crystal structure; inputting the atomic characteristics corresponding to the initial crystal structure into an initial neural network model to obtain a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted; determining model loss based on a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted, and adjusting model parameters of the initial neural network model based on the model loss until a model stopping training condition is met, so as to obtain a crystal structure prediction model.
As shown in fig. 3, the method includes:
s310, acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted.
S320, extracting features of the initial crystal structure of the material to be predicted, and obtaining atomic features corresponding to the initial crystal structure.
S330, inputting the atomic characteristics corresponding to the initial crystal structure into an initial neural network model to obtain a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted.
S340, determining model loss based on the crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted, and adjusting model parameters of the initial neural network model based on the model loss until the model stopping training condition is met, so as to obtain a crystal structure prediction model.
S350, performing crystal parameter optimization processing on the crystal structure prediction model to obtain target crystal structure parameters.
The initial neural network model may be, for example, a recurrent neural network. Specifically, feature extraction can be performed on an initial crystal structure of a material to be predicted in training data to obtain atomic features corresponding to atoms in the initial crystal structure, then the atomic features corresponding to the initial crystal structure can be input into a cyclic neural network to obtain a crystal energy prediction result, and further model loss is determined based on the crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted, wherein the loss function can be a mean square error loss function, and further model parameters of the cyclic neural network are updated based on the model loss until model training stop conditions are met, training is stopped, and a crystal structure prediction model is generated and stored.
Optionally, extracting features of the initial crystal structure of the material to be predicted to obtain atomic features corresponding to the initial crystal structure, including: atomic path extraction is carried out on an initial crystal structure of a material to be predicted, and at least one nearest neighbor path is obtained; and extracting atomic characteristics of each atom in the nearest neighbor path to obtain atomic characteristics corresponding to the initial crystal structure.
Specifically, the nearest neighbor atoms of each atom can be obtained according to the arrangement sequence of the atoms in the crystal structure, so that a nearest neighbor path from each atom is formed; further, atomic feature extraction can be performed on each atom in the nearest neighbor path through a feature extraction function, so that atomic features corresponding to the initial crystal structure are obtained.
Optionally, atomic path extraction is performed on an initial crystal structure of the material to be predicted to obtain at least one nearest neighbor path, including: repeating the following steps for each atom in the initial crystal structure of the material to be predicted: determining nearest neighbor atoms of current atoms in an initial crystal structure of the material to be predicted, and taking the nearest neighbor atoms of the current atoms as next extraction atoms; and determining the nearest neighbor atom of the next extraction atom until the next extraction atom meets the atomic path extraction stopping condition, and obtaining the nearest neighbor path.
The atomic path extraction stop condition may be that no atom appears within a truncated radius of the next extraction atom.
Illustratively, the atom R is selected from an initial crystal structure 1 Further determining the cutting radius R c Internal and atomic R 1 Nearest neighbor atom R 2 Further, the atom R is calculated 2 In order to avoid the circulation path, the nearest neighbor atoms of the n nearest neighbor paths P= { P can be obtained by repeating the above steps without repeating the atom extraction 1 ,P 2 …P n }。
Alternatively, the feature extraction function may include G 2 And G 4 Wherein G is 2 The formula of (2) is specifically:
G 4 the formula of (2) is specifically:
wherein R is i Represents the ith atom, R ij Represents the distance between atom i and atom j, η represents the width of the gaussian function, R s Represents the distance moved by the center of the Gaussian function, R c Represents the radius of the cut-off, θ ijk ＝acos(R ij ·R ik /R ij ·R ik )，Represents the angular resolution, λε [ -1,1 []。
Optionally, in the training process of the recurrent neural network, model weights can be calculated according to the distances between atoms, so that the recurrent neural network can sense the distances between the atoms, and the prediction performance of model prediction on the crystal structure is improved.
According to the technical scheme, the atomic characteristics corresponding to the initial crystal structure are obtained through feature extraction of the initial crystal structure of the material to be predicted, then the atomic characteristics corresponding to the initial crystal structure are input into the initial neural network model, the crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted is obtained, further model loss is determined based on the crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted, model parameters of the initial neural network model are adjusted based on the model loss until model stop training conditions are met, crystal structure prediction model training is achieved, and a reliable crystal structure prediction model is provided for crystal parameter optimization.
Example IV
Fig. 4 is a flowchart of a crystal structure prediction method according to a fourth embodiment of the present invention, where the method according to the present embodiment may be combined with each of the alternatives in the crystal structure prediction method according to the foregoing embodiment. The crystal structure prediction method provided in this embodiment is further optimized. Optionally, performing crystal parameter optimization processing on the crystal structure prediction model to obtain a target crystal structure parameter, including: determining a crystal structure parameter in the case that the predicted energy of the crystal structure prediction model is a minimum value; and taking the crystal structure parameter under the condition that the prediction energy of the crystal structure prediction model is the minimum value as the target crystal structure parameter.
As shown in fig. 4, the method includes:
s410, acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted.
S420, training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model.
S430, determining a crystal structure parameter under the condition that the prediction energy of the crystal structure prediction model is the minimum value.
S440, taking the crystal structure parameter under the condition that the prediction energy of the crystal structure prediction model is the minimum value as a target crystal structure parameter.
In this embodiment, the crystal structure parameter under the condition that the predicted energy of the crystal structure prediction model is the minimum value may be determined by a global optimization algorithm, which may be a bayesian optimization algorithm, a genetic algorithm, or the like, which is not limited herein.
For example, the crystal structure parameters under the condition that the predicted energy of the crystal structure prediction model is the minimum value can be determined through a bayesian optimization algorithm, and a specific solving formula is as follows:
x * ＝argmin x∈X f(x)
wherein f (x) represents a crystal structure prediction model; x represents a set of set values of the hyper-parameters, which may include crystal constant, axis angle, and atomic coordinates within the crystal; x represents a hyper-parameter set; x is x * Representing the target crystal structure parameter. It should be noted that, the process of solving the target crystal structure parameter by using the bayesian optimization algorithm may be divided into learning a proxy model and determining to output the next acquisition point through an acquisition function. It should be noted that, by introducing the acquisition function, the optimal solution of the crystal structure prediction model can be obtained with a small number of evaluation times.
According to the technical scheme provided by the embodiment of the invention, the crystal structure parameter under the condition that the output of the crystal structure prediction model is the minimum value is determined, and then the crystal structure parameter under the condition that the output of the crystal structure prediction model is the minimum value is taken as the target crystal structure parameter, so that the global optimal solution of the crystal structure parameter is obtained, and the accuracy of predicting the crystal structure is improved.
Example five
Fig. 5 is a flowchart of a crystal structure prediction method according to a fifth embodiment of the present invention, where the method according to the present embodiment may be combined with each of the alternatives in the crystal structure prediction method according to the foregoing embodiment. The crystal structure prediction method provided in this embodiment is further optimized. Optionally, the number of target crystal structure parameters is a plurality; correspondingly, after the crystal parameter optimization processing is performed on the crystal structure prediction model to obtain the target crystal structure parameter, the method further comprises the following steps: and sequencing the predicted energy of the corresponding structure of each target crystal structure parameter, and determining a recommended crystal structure based on the sequencing result.
As shown in fig. 5, the method includes:
s510, acquiring an initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted.
S520, training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model.
S530, performing crystal parameter optimization processing on the crystal structure prediction model to obtain a plurality of target crystal structure parameters.
S540, sorting the predicted energy of the corresponding structure of each target crystal structure parameter, and determining a recommended crystal structure based on the sorting result.
After obtaining a plurality of target crystal structure parameters, the predicted energy of the structure corresponding to the target crystal structure parameters can be calculated through VASP, then the crystal structures are ordered according to the predicted energy, and the N crystal structures with the largest or smallest crystal structures in the ordered result are output as recommended crystal structures, so that the screening of the crystal structures is realized. Wherein, N can be set up by user according to the prediction requirement, and is not limited herein.
Optionally, after determining the recommended crystal structure based on the ordering result, the method further comprises: under the condition that the recommended crystal structure does not meet the potential energy condition, adding the recommended crystal structure which does not meet the potential energy condition and an energy label corresponding to the recommended crystal structure which does not meet the potential energy condition into a training data set; training the initial neural network model based on the updated training data set to obtain an optimized crystal structure prediction model; and carrying out crystal parameter optimization treatment on the optimized crystal structure prediction model to obtain optimized target crystal structure parameters.
The optimized crystal structure prediction model refers to a crystal structure prediction model with updated model parameters.
Fig. 6 is a flowchart of a crystal structure prediction method according to an embodiment of the present invention. Specifically, component information and proportioning information of a material to be predicted can be input through input equipment of electronic equipment, and then various initial crystal structures of the material to be predicted are generated by widely sampling according to the component information and the proportioning information of the material to be predicted, and energy determination is carried out on each initial crystal structure of the material to be predicted, so that energy labels corresponding to each initial crystal structure of the material to be predicted are obtained; in addition, in the process of generating the training data, in order to reduce the time for generating the training data, when the initial crystal structure is generated, the crystal structure of which the same proportion constituent components are smaller than the initial crystal structure can be generated, so that the time is reduced while the training samples are enriched. Further, training data can be used for training a cyclic neural network integrated with distance weight perception to obtain an optimized crystal structure prediction model; and carrying out crystal parameter optimization processing on the optimized crystal structure prediction model to obtain a plurality of target crystal structure parameters, further calculating the prediction energy of the structure corresponding to the target crystal structure parameters through VASP, further sequencing the crystal structures according to the prediction energy, and outputting TOP-N crystal structures in the sequencing result as recommended crystal structures. Further, if the recommended crystal structure does not meet the potential energy condition, that is, the difference between the potential energy obtained by calculating the recommended crystal structure through VASP and the potential energy obtained by predicting the crystal structure by using the crystal structure prediction model is greater than a preset potential energy threshold, the recommended crystal structure and the potential energy obtained by calculating the recommended crystal structure through VASP can be added into the training dataset, and then the model is updated in a mode of performing incremental learning or relearning based on the updated training dataset until the predicted crystal structure meets the potential energy condition, and iteration is stopped.
Example six
Fig. 7 is a schematic structural diagram of a crystal structure prediction apparatus according to a sixth embodiment of the present invention. As shown in fig. 7, the apparatus includes:
the training data acquisition module 610 is configured to acquire an initial crystal structure of a material to be predicted and an energy tag corresponding to the initial crystal structure of the material to be predicted;
the prediction model training module 620 is configured to train an initial neural network model based on an initial crystal structure of a material to be predicted and an energy tag corresponding to the initial crystal structure of the material to be predicted, to obtain a crystal structure prediction model;
and the crystal structure parameter acquisition module 630 is configured to perform crystal parameter optimization processing on the crystal structure prediction model to obtain a target crystal structure parameter.
According to the technical scheme, the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted are obtained, and then the initial neural network model is trained based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted, so that a crystal structure prediction model is obtained, and crystal parameter optimization processing is carried out on the crystal structure prediction model, so that target crystal structure parameters are obtained. According to the technical scheme, the neural network prediction model and the parameter optimization processing method are combined to predict the crystal structure, so that the global optimal solution of the crystal structure is obtained, and compared with the traditional crystal structure prediction method, the crystal structure prediction operation time is shortened, and therefore the efficiency of predicting the crystal structure is improved.
In some alternative embodiments, the training data acquisition module 610 is further configured to:
acquiring component information and proportioning information of a material to be predicted;
generating a plurality of initial crystal structures of the material to be predicted based on the component information and the proportioning information of the material to be predicted;
and determining energy of each initial crystal structure of the material to be predicted, and obtaining energy labels corresponding to each initial crystal structure of the material to be predicted.
In some alternative embodiments, predictive model training module 620 includes:
the feature extraction unit is used for extracting features of the initial crystal structure of the material to be predicted to obtain atomic features corresponding to the initial crystal structure;
the crystal structure prediction unit is used for inputting the atomic characteristics corresponding to the initial crystal structure into an initial neural network model to obtain a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted;
and the prediction model training unit is used for determining model loss based on a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted, and adjusting model parameters of the initial neural network model based on the model loss until a model stopping training condition is met, so as to obtain a crystal structure prediction model.
In some alternative embodiments, the feature extraction unit includes:
the atomic path extraction unit is used for extracting an atomic path from the initial crystal structure of the material to be predicted to obtain at least one nearest neighbor path;
and the atomic characteristic extraction unit is used for extracting atomic characteristics of each atom in the nearest neighbor path to obtain atomic characteristics corresponding to the initial crystal structure.
In some alternative embodiments, the atom path extraction unit is specifically configured to:
repeating the following steps for each atom in the initial crystal structure of the material to be predicted:
determining nearest neighbor atoms of current atoms in the initial crystal structure of the material to be predicted, and taking the nearest neighbor atoms of the current atoms as next extraction atoms;
and determining the nearest neighbor atom of the next extraction atom until the next extraction atom meets the atomic path extraction stopping condition, and obtaining the nearest neighbor path.
In some alternative embodiments, the crystal structure parameter acquisition module 630 is further configured to:
determining a crystal structure parameter in the case that the predicted energy of the crystal structure prediction model is a minimum value;
and taking the crystal structure parameter under the condition that the prediction energy of the crystal structure prediction model is the minimum value as a target crystal structure parameter.
In some alternative embodiments, the number of target crystal structure parameters is a plurality;
correspondingly, the device further comprises:
and the recommended crystal structure determining module is used for sequencing the predicted energy of the structures corresponding to the target crystal structure parameters and determining the recommended crystal structure based on the sequencing result.
In some alternative embodiments, the apparatus further comprises:
the training data set updating module is used for adding the recommended crystal structure which does not meet the potential energy condition and the energy label corresponding to the recommended crystal structure which does not meet the potential energy condition into the training data set under the condition that the recommended crystal structure does not meet the potential energy condition;
the model updating module is used for training the initial neural network model based on the updated training data set to obtain an optimized crystal structure prediction model;
and the target crystal structure parameter optimization module is used for carrying out crystal parameter optimization processing on the optimized crystal structure prediction model to obtain optimized target crystal structure parameters.
The crystal structure prediction device provided by the embodiment of the invention can execute the crystal structure prediction method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Example seven
Fig. 8 shows a schematic diagram of the structure of an electronic device 10 that may be used to implement an embodiment of the invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices (e.g., helmets, eyeglasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 8, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An I/O interface 15 is also connected to bus 14.
Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as a crystal structure prediction method, which includes:
acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted;
Training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model;
and carrying out crystal parameter optimization treatment on the crystal structure prediction model to obtain target crystal structure parameters.
In some embodiments, the method of predicting a crystal structure may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the above-described crystal structure prediction method may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the method of predicting the crystal structure in any other suitable manner (e.g., by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (11)
1. A method for predicting a crystal structure, comprising:
acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted;
training an initial neural network model based on an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model;
and carrying out crystal parameter optimization treatment on the crystal structure prediction model to obtain target crystal structure parameters.
2. The method according to claim 1, wherein the obtaining the initial crystal structure of the material to be predicted and the energy tag corresponding to the initial crystal structure of the material to be predicted comprises:
acquiring component information and proportioning information of a material to be predicted;
generating a plurality of initial crystal structures of the material to be predicted based on the component information and the proportioning information of the material to be predicted;
and determining energy of each initial crystal structure of the material to be predicted, and obtaining energy labels corresponding to each initial crystal structure of the material to be predicted.
3. The method according to claim 1, wherein training the initial neural network model based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted to obtain the crystal structure prediction model includes:
Extracting features of the initial crystal structure of the material to be predicted to obtain atomic features corresponding to the initial crystal structure;
inputting the atomic characteristics corresponding to the initial crystal structure into an initial neural network model to obtain a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted;
determining model loss based on a crystal energy prediction result corresponding to the initial crystal structure of the material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted, and adjusting model parameters of the initial neural network model based on the model loss until a model stopping training condition is met, so as to obtain a crystal structure prediction model.
4. A method according to claim 3, wherein the feature extraction of the initial crystal structure of the material to be predicted to obtain atomic features corresponding to the initial crystal structure includes:
atomic path extraction is carried out on the initial crystal structure of the material to be predicted, so that at least one nearest neighbor path is obtained;
and extracting atomic characteristics of each atom in the nearest neighbor path to obtain atomic characteristics corresponding to the initial crystal structure.
5. The method of claim 4, wherein the performing atomic path extraction on the initial crystal structure of the material to be predicted to obtain at least one nearest neighbor path comprises:
repeating the following steps for each atom in the initial crystal structure of the material to be predicted:
determining nearest neighbor atoms of current atoms in the initial crystal structure of the material to be predicted, and taking the nearest neighbor atoms of the current atoms as next extraction atoms;
and determining the nearest neighbor atom of the next extraction atom until the next extraction atom meets the atomic path extraction stopping condition, and obtaining the nearest neighbor path.
6. The method of claim 1, wherein performing a crystal parameter optimization process on the crystal structure prediction model to obtain a target crystal structure parameter comprises:
determining a crystal structure parameter in the case that the predicted energy of the crystal structure prediction model is a minimum value;
and taking the crystal structure parameter under the condition that the prediction energy of the crystal structure prediction model is the minimum value as a target crystal structure parameter.
7. The method of claim 1, wherein the number of target crystal structure parameters is a plurality;
Correspondingly, after performing crystal parameter optimization processing on the crystal structure prediction model to obtain a target crystal structure parameter, the method further comprises:
and sequencing the predicted energy of the corresponding structure of each target crystal structure parameter, and determining a recommended crystal structure based on the sequencing result.
8. The method of claim 7, wherein after the determining of the recommended crystal structure based on the ordering result, the method further comprises:
under the condition that the recommended crystal structure does not meet the potential energy condition, adding the recommended crystal structure which does not meet the potential energy condition and an energy label corresponding to the recommended crystal structure which does not meet the potential energy condition into a training data set;
training the initial neural network model based on the updated training data set to obtain an optimized crystal structure prediction model;
and carrying out crystal parameter optimization processing on the optimized crystal structure prediction model to obtain optimized target crystal structure parameters.
9. A crystal structure prediction apparatus, comprising:
the training data acquisition module is used for acquiring an initial crystal structure of a material to be predicted and an energy label corresponding to the initial crystal structure of the material to be predicted;
The prediction model training module is used for training the initial neural network model based on the initial crystal structure of the material to be predicted and the energy label corresponding to the initial crystal structure of the material to be predicted to obtain a crystal structure prediction model;
and the crystal structure parameter acquisition module is used for carrying out crystal parameter optimization processing on the crystal structure prediction model to obtain target crystal structure parameters.
10. An electronic device, the electronic device comprising:
at least one processor;
and a memory communicatively coupled to the at least one processor;
wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of predicting a crystal structure of any one of claims 1-8.
11. A computer readable storage medium storing computer instructions for causing a processor to perform the method of predicting a crystal structure according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310686934.4A CN116665823A (en) | 2023-06-09 | 2023-06-09 | Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310686934.4A CN116665823A (en) | 2023-06-09 | 2023-06-09 | Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116665823A true CN116665823A (en) | 2023-08-29 |
Family
ID=87720468
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310686934.4A Pending CN116665823A (en) | 2023-06-09 | 2023-06-09 | Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116665823A (en) |
-
2023
- 2023-06-09 CN CN202310686934.4A patent/CN116665823A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113705628B (en) | Determination method and device of pre-training model, electronic equipment and storage medium | |
| CN116307215A (en) | Load prediction method, device, equipment and storage medium of power system | |
| CN113947188A (en) | Training method of target detection network and vehicle detection method | |
| CN112966744A (en) | Model training method, image processing method, device and electronic equipment | |
| CN116071608B (en) | Target detection method, device, equipment and storage medium | |
| CN115293149A (en) | Entity relationship identification method, device, equipment and storage medium | |
| CN116629620A (en) | Risk level determining method and device, electronic equipment and storage medium | |
| CN118210670A (en) | Log abnormality detection method and device, electronic equipment and storage medium | |
| CN116665823A (en) | Crystal structure prediction method, crystal structure prediction device, electronic equipment and storage medium | |
| CN117370914A (en) | Power grid attack detection model test method, device, equipment and storage medium | |
| CN116796282A (en) | Molecular screening method, training device, electronic equipment and storage medium | |
| CN114610953A (en) | Data classification method, device, equipment and storage medium | |
| CN114611609A (en) | Graph network model node classification method, device, equipment and storage medium | |
| CN113326885A (en) | Method and device for training classification model and data classification | |
| CN116933896B (en) | Super-parameter determination and semantic conversion method, device, equipment and medium | |
| CN118245823B (en) | Insulator leakage current prediction method, device, equipment and storage medium | |
| CN114581751B (en) | Training method of image recognition model, image recognition method and device | |
| CN117934137A (en) | Bad asset recovery prediction method, device and equipment based on model fusion | |
| CN114037057B (en) | Pre-training model generation method and device, electronic equipment and storage medium | |
| CN116111562A (en) | Method, device, equipment and storage medium for determining reliability of power distribution network | |
| CN117251809A (en) | Power grid time sequence data anomaly detection method, device, equipment and storage medium | |
| CN116913406A (en) | Transformer fault identification method, device, equipment and medium | |
| CN116307098A (en) | Model training method, insulator surface salt density prediction method and insulator surface salt density prediction device | |
| CN116434835A (en) | Protein structure identification method, training method and device | |
| CN117851598A (en) | Auxiliary classification method, device, equipment and medium for large language model |
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 |