Crystal structure prediction model training method and device, crystal structure prediction method and device, equipment and medium

By performing crystallization processing and coding module processing on the ray diffraction spectrum, a crystal structure prediction model is generated, which solves the problems of low accuracy and long time-consuming prediction of crystal structures in the prior art, and achieves fast and accurate prediction of material crystal structures.

CN119007883APending Publication Date: 2024-11-22SUZHOU INST FOR ADVANCED STUDY USTC +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411022110.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing crystal structure measurement methods have low accuracy and are time-consuming, making it difficult to quickly and accurately predict the crystal structure of the material.

Method used

By obtaining the ray diffraction spectrum of the target crystal, the crystal surfaceization module of the crystal structure prediction model is used to perform crystal surfaceization processing, and the crystal surface characteristics are generated, and inputting them into the encoding module for processing, to generate the target crystal structure prediction model for fast and accurate Predict the crystal structure of the material.

Benefits of technology

It improves the accuracy of crystal structure prediction, shortens the prediction time, and provides a fast and accurate method for predicting crystal structures of materials.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119007883A_ABST
    Figure CN119007883A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a crystal structure prediction model training method and device, a prediction method and device, equipment and a medium. The training method of the crystal structure prediction model comprises the following steps: acquiring a target crystal set, and determining a ray diffraction spectrum of each crystal; performing crystal face processing on each ray diffraction spectrum through a crystal face module of the crystal structure prediction model to obtain crystal face characteristics matched with each ray diffraction spectrum; and sequentially inputting each crystal face feature into a coding module of a crystal structure prediction model for processing, and generating a target crystal structure prediction model. According to the scheme provided by the embodiment of the invention, the problems of low prediction accuracy and long time consumption of the crystal structure are solved, and a basis is provided for rapidly and accurately predicting the crystal structure of the material.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a crystal structure prediction model training, prediction method, device, equipment and medium. Background Art

[0002] In materials science, the properties of a material are largely determined by its atomic arrangement and crystal structure. Determining a material's crystal structure is crucial to understanding its mechanical, electromagnetic, and thermodynamic properties.

[0003] Existing methods for crystal structure determination require a lot of manual work to obtain relevant background knowledge from the materials. This method has low accuracy and is time-consuming.

[0004] How to quickly and accurately predict the crystal structure of materials is a key research issue in the industry. Summary of the invention

[0005] The present invention provides a crystal structure prediction model training, prediction method, device, equipment and medium to solve the problem of low crystal structure prediction accuracy and long time consumption, and provide a basis for quickly and accurately predicting the crystal structure of a material.

[0006] According to one aspect of the present invention, a method for training a crystal structure prediction model is provided, the method comprising:

[0007] Acquire a target crystal set and determine the X-ray diffraction spectrum of each crystal;

[0008] Performing crystal plane processing on each of the ray diffraction spectra respectively through a crystal plane module of a crystal structure prediction model to obtain crystal plane features matching each of the ray diffraction spectra;

[0009] The crystal plane features are sequentially input into the encoding module of the crystal structure prediction model for processing to generate a target crystal structure prediction model.

[0010] According to another aspect of the present invention, a method for predicting a crystal structure is provided, the method comprising:

[0011] Obtaining a X-ray diffraction spectrum of a target crystal;

[0012] Inputting the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal;

[0013] Wherein, the crystal structure prediction model is trained by the crystal structure prediction model training method described in any embodiment of the present invention.

[0014] According to another aspect of the present invention, there is provided a training device for a crystal structure prediction model, the device comprising:

[0015] An acquisition module is used to acquire a target crystal set and determine the ray diffraction spectrum of each crystal;

[0016] A crystal plane feature determination module, used to perform crystal plane processing on each of the ray diffraction spectra through a crystal plane module of a crystal structure prediction model, so as to obtain crystal plane features matching each of the ray diffraction spectra;

[0017] The model generation module is used to input each of the crystal plane features into the encoding module of the crystal structure prediction model in sequence for processing to generate a target crystal structure prediction model.

[0018] According to another aspect of the present invention, there is provided a device for predicting a crystal structure, the device comprising:

[0019] An acquisition module, used to acquire the ray diffraction spectrum of the target crystal;

[0020] A crystal structure determination module, used for inputting the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal;

[0021] Wherein, the crystal structure prediction model is trained by the crystal structure prediction model training method described in any embodiment of the present invention.

[0022] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0023] at least one processor; and

[0024] a memory communicatively connected to the at least one processor; wherein,

[0025] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the training method of the crystal structure prediction model or the crystal structure prediction method described in any embodiment of the present invention.

[0026] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the training method of a crystal structure prediction model or the crystal structure prediction method described in any embodiment of the present invention when executed.

[0027] The technical solution of the embodiment of the present invention obtains a target crystal set and determines the ray diffraction spectrum of each crystal; performs crystal faceting on each of the ray diffraction spectra respectively through a crystal faceting module of a crystal structure prediction model to obtain crystal face features matching each of the ray diffraction spectra; and sequentially inputs each of the crystal face features into an encoding module of the crystal structure prediction model for processing to generate a target crystal structure prediction model, thereby solving the problem of low accuracy and long time consumption in crystal structure prediction and providing a basis for quickly and accurately predicting the crystal structure of a material.

[0028] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0030] Figure 1 is a flow chart of a method for training a crystal structure prediction model provided according to Embodiment 1 of the present invention;

[0031] Figure 2 is a schematic diagram of a diffraction process provided according to Embodiment 1 of the present invention;

[0032] Figure 3 is a schematic diagram of the structure of a crystal structure prediction model provided according to Embodiment 1 of the present invention;

[0033] Figure 4 is a flow chart of a method for predicting a crystal structure provided according to Embodiment 2 of the present invention;

[0034] Figure 5 is a schematic structural diagram of a crystal structure prediction model training device provided according to Embodiment 3 of the present invention;

[0035] Figure 6 is a schematic diagram of the structure of a crystal structure prediction device provided according to a fourth embodiment of the present invention;

[0036] Figure 7 It is a structural schematic diagram of an electronic device for implementing a crystal structure prediction model training method or a crystal structure prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0038] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0039] Embodiment 1

[0040] Figure 1 This is a flow chart of a method for training a crystal structure prediction model according to the first embodiment of the present invention. This embodiment is applicable to the case of training a neural network model for predicting crystal structure. The method can be executed by a training device for a crystal structure prediction model. The training device for a crystal structure prediction model can be implemented in the form of hardware and / or software. The training device for a crystal structure prediction model can be configured in an electronic device such as a computer or a server. Figure 1 As shown, the method includes:

[0041] Step 110: Acquire a target crystal set and determine the ray diffraction spectrum of each crystal.

[0042] The target crystal set includes a plurality of different types of crystal materials, such as metal crystal materials, semiconductor crystal materials or optical crystal materials, etc., which are not limited in this embodiment.

[0043] Among them, the ray diffraction spectrum of each crystal can be an X-ray diffraction spectrum (X-ray diffraction, Xrd); in an optional implementation method of this embodiment, X-ray diffraction can be performed on each crystal material in the target crystal concentration by X-rays to obtain the X-ray diffraction spectrum of each crystal.

[0044] Step 120: Perform crystal plane processing on each of the ray diffraction spectra respectively through the crystal plane module of the crystal structure prediction model to obtain crystal plane features matching each of the ray diffraction spectra.

[0045] The crystal structure prediction model involved in this embodiment may include a crystal faceting module and an encoding module; wherein the crystal faceting module can model the ray diffraction spectrum of the crystal material from a crystal face perspective, so as to obtain crystal face features that match it; the encoding module may include a state space model block as well as a convolution layer and a pooling layer, etc., which are not limited in this embodiment.

[0046] It should be noted that the X-ray diffraction spectrum of any crystalline material is a curve with a length of n, which can be expressed as S = [S1, S2, ..., S θ ,…,S n ],(θ∈[0°,δ,2δ,3δ,…,180°]); where θ represents the diffraction angle of X-ray intake, S θ It represents the diffraction intensity obtained when the incident angle is θ, and δ records the interval of diffraction intensity.

[0047] Usually, when analyzing the crystal structure, some key segments are intercepted for processing, because the diffraction intensity at high angles is very small and has little effect on analyzing the crystal structure. Therefore, the selection range of θ can be set to an interval R(θ1,θ r ,δ), where 0°≤θ1≤θ r ≤180°, usually, θ1≤5°, 50°≤θ r ≤130°, R(θ1,θ r ,δ) means [θ1, θ1+δ, θ1+2δ,…,θ r ].

[0048] Figure 2 is a schematic diagram of a diffraction process according to the first embodiment of the present invention, such as Figure 2 As shown, in the diffraction process, each diffraction angle corresponds to a crystal plane in the crystal. The diffraction intensity at this diffraction angle represents the number of such crystal planes, that is, the higher the intensity, the greater the number of such crystal planes. That is, the position of each peak in the X-ray diffraction spectrum indicates the existence of important crystal planes in the crystal; it can be understood that these crystal planes rather than the intensity reflect the structural information to a large extent, and the intensity of each peak in the X-ray diffraction spectrum only reflects the importance of each crystal plane in the crystal. Based on this, in this embodiment, R(θ1,θ r , δ) represents a crystal plane, and S θ To indicate the importance of this crystal face in the crystal, the X-ray diffraction spectrum data of a crystal is divided into two sequences S and Process separately.

[0049] In an optional implementation of the present embodiment, each of the ray diffraction spectra is subjected to crystallization processing respectively through a crystallization module of a crystal structure prediction model to obtain crystal plane features matching each of the ray diffraction spectra, which may include: determining a crystal plane vector matching a target ray diffraction spectrum, and respectively determining the peak intensity of each of the crystal planes; the crystal plane vectors are used to characterize different crystal planes in the target ray diffraction spectrum; and obtaining crystal plane features matching the target ray diffraction spectrum according to the crystal plane vectors and the peak intensity of each of the crystal planes.

[0050] The target X-ray diffraction spectrum may be an X-ray diffraction spectrum of any crystal in the target crystal set, which is not limited in this embodiment.

[0051] Optionally, in this embodiment, the crystallization module can determine the crystal plane vector that matches the target ray diffraction spectrum, and determine the peak intensity of each crystal plane. Further, the crystal plane characteristics that match the target ray diffraction spectrum can be obtained based on the determined crystal plane vector and the peak intensity of each crystal plane.

[0052] In an optional implementation of this embodiment, determining the crystal plane vector that matches the target ray diffraction spectrum may include: discretizing the target ray diffraction spectrum to obtain a discrete target ray diffraction spectrum; and mapping the discrete target ray diffraction spectrum to obtain the crystal plane vector.

[0053] In a specific implementation, the target ray diffraction spectrum can be discretized first, that is, each crystal plane in the target ray diffraction spectrum can be tokenized to obtain a discrete target ray diffraction spectrum; further, the discrete target ray diffraction spectrum can be mapped. For example, each θ diffraction angle can be embedded into a k-dimensional vector through a linear projection layer, that is, an n× crystal plane vector is obtained to respectively characterize n different crystal planes.

[0054] Furthermore, the peak intensity S at each diffraction angle θ is θ It is regarded as a description of the importance of each crystal plane and is used to weight the crystal plane characterization; that is, the peak intensity vector and the crystal plane vector are multiplied to obtain an n×k crystal plane feature S'; where,

[0055] Step 130: input each of the crystal plane features into the encoding module of the crystal structure prediction model in sequence for processing to generate a target crystal structure prediction model.

[0056] In an optional implementation of the present embodiment, after obtaining the crystal plane features matching each ray diffraction spectrum through the crystal plane module of the crystal structure prediction model, these crystal plane features can be further input into the encoding module of the crystal structure prediction model for processing, and the final target crystal structure prediction model can be generated when the iteration stop condition is met; in the present embodiment, the trained crystal structure prediction model can be used to predict the crystal structure of any crystal.

[0057] In this embodiment, the encoding module of the crystal structure prediction model may include multiple state space model blocks, multiple convolution and pooling layers, which are not limited in this embodiment; illustratively, the state space model block may be a Mamba block; it should be noted that the Mamba model is a sequence transformation model, which effectively captures the dependency information in a long sequence by adding a dependent data selection mechanism and a hardware-aware parallel algorithm to the state space model S4, and maintains an almost linear ultra-high computational efficiency. The Mamba model consists of multiple stacked Mamba blocks, each of which converts a sequence (where B represents the batch size, L represents the sequence length, and D represents the dimension of each item in the sequence) as input. In each Mamba block, the data undergoes operations such as projection, state space model, and residual link, and finally obtains an output sequence

[0058] Optionally, in this embodiment, each of the crystal plane features is sequentially input into the encoding module of the crystal structure prediction model for processing, which may include: performing vector transformation on the crystal plane features through at least one state space model block to determine each target crystal plane and context sequence information matching each target crystal plane; processing each of the target crystal planes and each of the context sequence information through multiple residual convolution blocks to obtain a crystal representation matching the target crystal.

[0059] In a specific implementation, the crystal plane features can be transformed by seq2seq through multiple Mamba blocks (for example, 100, 2000 or 10000, etc., which are not limited in this embodiment), so as to obtain each crystal plane and the context sequence information matching each crystal plane; further, further characterization can be performed through multiple stacked one-dimensional residual convolution blocks and pooling layers. Among them, the residual convolution block is mainly composed of residual connection, one-dimensional convolution and regularization module; wherein, the kernels in the one-dimensional convolution layer and the pooling layer are small kernels, such as 1×1, and the number of channels of the convolution block gradually increases; in this embodiment, by repeatedly stacking multiple small kernel layers and reducing the dimension layer by layer, the information of all crystal planes in the X-ray diffraction spectrum can be gathered, thereby obtaining the characterization of the crystal.

[0060] In an optional implementation of this embodiment, after obtaining the crystal representation matching the target crystal, the method may further include: classifying the crystal representation to determine the crystal structure matching the target crystal.

[0061] Optionally, in this embodiment, after obtaining the crystal features that match the target crystal according to the encoding module of the crystal structure prediction model, the crystal features can be further input into the classifier of the crystal structure prediction model, and then the crystal representations can be classified by the classifier to determine the crystal structure that matches the crystal.

[0062] In order to better understand the crystal structure prediction model involved in this embodiment, Figure 3 is a schematic diagram of a crystal structure prediction model provided according to the first embodiment of the present invention. Figure 3 As shown, it mainly includes a crystallization module 310 , an encoding module 320 and a classifier 330 .

[0063] In an example of this embodiment, during the training process, the number of Mamba blocks can be set to 8, the dimension of the crystal plane representation can be set to 8, and the hidden layer dimension of the state space model can be set to 16. The Adam optimizer with a weight decay of 1e-5 and a learning rate of 5e-5 is used to optimize the model, and the learning rate is adjusted using a cosine curve learning rate scheduler. Finally, the crystal structure prediction model can reach full convergence within 100 layers.

[0064] The technical solution of this embodiment is to obtain a target crystal set and determine the ray diffraction spectrum of each crystal; perform crystal faceting on each of the ray diffraction spectra respectively through the crystal faceting module of the crystal structure prediction model to obtain crystal face features matching each of the ray diffraction spectra; input each of the crystal face features into the encoding module of the crystal structure prediction model in turn for processing to generate a target crystal structure prediction model, thereby solving the problem of low accuracy and long time consumption in crystal structure prediction and providing a basis for quickly and accurately predicting the crystal structure of a material.

[0065] Embodiment 2

[0066] Figure 4 This is a flow chart of a method for predicting a crystal structure according to the second embodiment of the present invention. This embodiment is applicable to the case where the crystal structure prediction model obtained by training in the above embodiment is used to predict the crystal structure of any crystal. The method can be executed by a crystal structure prediction device, which can be implemented in the form of hardware and / or software. The crystal structure prediction device can be configured in electronic devices such as computers or servers. Figure 4 As shown, the method includes:

[0067] Step 410: Obtain the X-ray diffraction spectrum of the target crystal.

[0068] Among them, the target crystal can be any crystal material with a determined crystal structure, which may not be included in the target crystal set involved in the above embodiments; in this embodiment, the target crystal can be subjected to X-ray diffraction treatment in the laboratory to obtain the X-ray diffraction spectrum of the crystal.

[0069] Step 420: input the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal.

[0070] The crystal structure prediction model is obtained by training using the crystal structure prediction model training method involved in any of the above embodiments.

[0071] In an optional implementation of the present embodiment, after obtaining the X-ray diffraction spectrum of the target crystal, the X-ray diffraction spectrum can be input into the trained crystal structure prediction model for prediction, that is, the X-ray diffraction spectrum of the target crystal is processed in sequence by the crystal faceting module, encoding module and classifier of the trained crystal structure prediction model, so as to predict the space group to which the target crystal belongs, that is, the crystal structure of the target crystal.

[0072] The solution of this embodiment can analyze the X-ray diffraction spectrum of any crystal through the crystal structure prediction model obtained through training to predict its crystal structure, thereby solving the problem of low accuracy and long time consumption of crystal structure prediction, and can quickly and accurately predict the crystal structure of the material.

[0073] Embodiment 3

[0074] Figure 5 Schematic diagram of a crystal structure prediction model training device according to Embodiment 3 of the present invention. Figure 5 As shown, the device includes: an acquisition module 510 , a crystal plane feature determination module 520 and a model generation module 530 .

[0075] An acquisition module 510 is used to acquire a target crystal set and determine the ray diffraction spectrum of each crystal;

[0076] A crystal plane feature determination module 520, configured to perform crystal plane processing on each of the ray diffraction spectra respectively through a crystal plane module of a crystal structure prediction model, so as to obtain a crystal plane feature matching each of the ray diffraction spectra;

[0077] The model generation module 530 is used to input each of the crystal plane features into the encoding module of the crystal structure prediction model in sequence for processing to generate a target crystal structure prediction model.

[0078] The solution of this embodiment obtains the target crystal set through the acquisition module and determines the ray diffraction spectrum of each crystal; performs crystal face processing on each of the ray diffraction spectra through the crystal face feature determination module to obtain the crystal face features matching each of the ray diffraction spectra; and inputs each of the crystal face features into the encoding module of the crystal structure prediction model in sequence through the model generation module for processing to generate the target crystal structure prediction model, thereby solving the problem of low accuracy and long time consumption in crystal structure prediction and providing a basis for quickly and accurately predicting the crystal structure of the material.

[0079] In an optional implementation of this embodiment, the crystal plane feature determination module 520 is specifically used to determine the crystal plane vectors matching the target ray diffraction spectrum, and respectively determine the peak intensity of each of the crystal planes; the crystal plane vectors are used to characterize different crystal planes in the target ray diffraction spectrum;

[0080] A crystal plane feature matching the target ray diffraction spectrum is obtained according to the crystal plane vector and the peak intensity of each crystal plane.

[0081] In an optional implementation of this embodiment, the crystal plane feature determination module 520 is further used to discretize the target ray diffraction spectrum to obtain a discrete target ray diffraction spectrum;

[0082] The discrete target ray diffraction spectrum is mapped to obtain the crystal plane vector.

[0083] In an optional implementation of this embodiment, the model generation module 530 is specifically used to perform vector transformation on the crystal plane features through at least one state space model block to determine each target crystal plane and context sequence information matching each target crystal plane;

[0084] Each of the target crystal planes and each of the context sequence information is processed through a plurality of residual convolution blocks to obtain a crystal representation matching the target crystal.

[0085] In an optional implementation of this embodiment, the training device for the crystal structure prediction model further includes: a classification module, which is used to classify the crystal representation and determine the crystal structure that matches the target crystal.

[0086] The training device for the crystal structure prediction model provided in the embodiment of the present invention can execute the training method for the crystal structure prediction model provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0087] Embodiment 4

[0088] Figure 6 Schematic diagram of a crystal structure prediction device provided according to the third embodiment of the present invention. Figure 6As shown, the device includes: an acquisition module 610 and a crystal structure determination module 620.

[0089] An acquisition module 610 is used to acquire a ray diffraction spectrum of a target crystal;

[0090] A crystal structure determination module 620 is used to input the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal;

[0091] Wherein, the crystal structure prediction model is trained by the crystal structure prediction model training method described in any of the above embodiments.

[0092] The solution of this embodiment obtains the X-ray diffraction spectrum of the target crystal through an acquisition module; inputs the X-ray diffraction spectrum into a pre-trained crystal structure prediction model through a crystal structure determination module to obtain the crystal structure of the target crystal, thereby solving the problem of low accuracy and long time consumption in crystal structure prediction, and can quickly and accurately predict the crystal structure of the material.

[0093] Embodiment 5

[0094] Figure 7 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0095] like Figure 7 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can 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 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0096] A number of 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, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0097] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as a training method for a crystal structure prediction model, or a method for predicting a crystal structure.

[0098] In some embodiments, the training method of the crystal structure prediction model, or the prediction method of the crystal structure can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, the training method of the crystal structure prediction model described above, or one or more steps of the prediction method of the crystal structure can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the training method of the crystal structure prediction model, or the prediction method of the crystal structure by any other appropriate means (for example, by means of firmware).

[0099] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0100] Computer programs for implementing the 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, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0101] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0102] To provide interaction with a user, the systems and techniques described herein may 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 the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the 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, voice input, or tactile input).

[0103] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0104] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0105] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0106] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for training a crystal structure prediction model, characterized in that: include: Acquire a target crystal set and determine the X-ray diffraction spectrum of each crystal; Performing crystal plane processing on each of the ray diffraction spectra respectively through a crystal plane module of a crystal structure prediction model to obtain crystal plane features matching each of the ray diffraction spectra; The crystal plane features are sequentially input into the encoding module of the crystal structure prediction model for processing to generate a target crystal structure prediction model.

2. The method for training a crystal structure prediction model according to claim 1, characterized in that: The crystallization module of the crystal structure prediction model performs crystallization processing on each of the ray diffraction spectra to obtain crystal plane features matching each of the ray diffraction spectra, including: Determine a crystal plane vector that matches the target ray diffraction spectrum, and respectively determine the peak intensity of each crystal plane; the crystal plane vector is used to characterize different crystal planes in the target ray diffraction spectrum; A crystal plane feature matching the target ray diffraction spectrum is obtained according to the crystal plane vector and the peak intensity of each crystal plane.

3. The method for training a crystal structure prediction model according to claim 2, characterized in that: The step of determining a crystal plane vector matching a target ray diffraction spectrum comprises: Discretizing the target ray diffraction spectrum to obtain a discrete target ray diffraction spectrum; The discrete target ray diffraction spectrum is mapped to obtain the crystal plane vector.

4. The method for training a crystal structure prediction model according to claim 1, characterized in that: The step of sequentially inputting the crystal plane features into the encoding module of the crystal structure prediction model for processing comprises: Performing vector transformation on the crystal plane features through at least one state space model block to determine each target crystal plane and context sequence information matching each target crystal plane; Each of the target crystal planes and each of the context sequence information is processed through a plurality of residual convolution blocks to obtain a crystal representation matching the target crystal.

5. The method for training a crystal structure prediction model according to claim 4, characterized in that: After obtaining a crystal characterization that matches the target crystal, it also includes: The crystal characterizations are classified and the crystal structure matching the target crystal is determined.

6. A method for predicting a crystal structure, characterized in that: include: Obtaining a X-ray diffraction spectrum of a target crystal; Inputting the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal; Wherein, the crystal structure prediction model is trained by the crystal structure prediction model training method described in any one of claims 1-5.

7. A training device for a crystal structure prediction model, characterized in that: include: An acquisition module is used to acquire a target crystal set and determine the ray diffraction spectrum of each crystal; A crystal plane feature determination module, used to perform crystal plane processing on each of the ray diffraction spectra through a crystal plane module of a crystal structure prediction model, so as to obtain crystal plane features matching each of the ray diffraction spectra; The model generation module is used to input each of the crystal plane features into the encoding module of the crystal structure prediction model in sequence for processing to generate a target crystal structure prediction model.

8. A crystal structure prediction device, characterized in that: include: An acquisition module, used to acquire the ray diffraction spectrum of the target crystal; A crystal structure determination module, used for inputting the X-ray diffraction spectrum into a pre-trained crystal structure prediction model to obtain the crystal structure of the target crystal; Wherein, the crystal structure prediction model is trained by the crystal structure prediction model training method described in any one of claims 1-5.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the training method of the crystal structure prediction model described in any one of claims 1 to 5, or the crystal structure prediction method described in claim 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the training method of a crystal structure prediction model described in any one of claims 1 to 5, or the crystal structure prediction method described in claim 6 when executed.

Citation Information

Cited By

  • Crystal structure prediction method and device, electronic equipment and storage medium

    CN119673349A

  • Crystal material generation method and system based on energy band condition and space group symmetry constraint

    CN120636634A