Crystal structure similarity calculation method, screening method, device and equipment

Abstract

The application relates to a crystal structure similarity calculation method, a screening method, a device and equipment. The method comprises the following steps: obtaining two crystal structure models; calculating a vector formed by each matching type of element pair in each crystal structure model; calculating a target value density distribution according to the vector formed by each matching type of element pair; and determining the similarity of the two crystal structure models according to the target value density distribution corresponding to the vector formed by each matching type of element pair in each crystal structure model. The scheme provided by the application can calculate the similarity between two crystal structures, reduce the computing resources, and shorten the calculation time.

Description

Technical Field

[0001] This application relates to the field of molecular dynamics simulation and calculation technology, and in particular to a method, screening method, apparatus and equipment for calculating crystal structure similarity. Background Technology

[0002] In the field of molecular dynamics simulations, it is often necessary to compare different crystal structures and calculate the similarity between two crystal structures for experimental analysis or to determine whether two crystal structures belong to the same type of structure.

[0003] In related technologies, the degree of difference between two crystal structures is measured by calculating the RMSD (Root Mean Square Deviation) between them and then using the RMSD output to calculate the similarity between the two crystal structures.

[0004] However, calculating RMSD is computationally intensive, time-consuming, and requires a large amount of computing resources, making it difficult to apply and popularize. Summary of the Invention

[0005] To address or partially address the problems existing in related technologies, this application provides a method, screening method, apparatus, and equipment for calculating crystal structure similarity, which can calculate the similarity between two crystal structures, reduce computational resources, and shorten computation time.

[0006] The first aspect of this application provides a method for calculating crystal structure similarity, including:

[0007] Obtain two crystal structure models;

[0008] Calculate the vector formed by each matching element pair in each of the crystal structure models; wherein the element pair includes two matching elements;

[0009] Calculate the target value density distribution based on the vector formed by the element pairs of each matching type;

[0010] The similarity between two crystal structure models is determined based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model.

[0011] In one implementation, the elements include atoms, groups of atoms, or molecules, and calculating the vector formed by element pairs of each matching type in each of the crystal structure models includes:

[0012] Obtain the center point of the elements contained in each matching type element pair in each crystal structure model;

[0013] The vector formed by the element pair is determined based on the center points of the two elements in the element pair;

[0014] The center point of the element includes the geometric center, centroid, or weighted geometric center of the element.

[0015] A second aspect of this application provides a method for screening crystal structures, comprising:

[0016] Obtain multiple crystal structures;

[0017] The similarity between any two crystal structures can be calculated using the crystal structure similarity calculation method described above.

[0018] Output the crystal structure whose similarity satisfies the preset conditions.

[0019] A third aspect of this application provides a crystal structure similarity calculation device, comprising:

[0020] The acquisition module is used to acquire two crystal structure models;

[0021] The first calculation module is used to calculate the vector formed by each matching element pair in the crystal structure model obtained by each acquisition module; wherein, the element pair includes two matching elements;

[0022] The second calculation module is used to calculate the target value density distribution based on the vector formed by the element pairs of each matching type calculated by the first calculation module.

[0023] The output module is used to determine the similarity between two crystal structure models based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model calculated by the second calculation module.

[0024] A fourth aspect of this application provides a crystal structure screening device, comprising:

[0025] The acquisition module is used to acquire multiple crystal structures;

[0026] The calculation module is used to calculate the similarity between any two crystal structures using the crystal structure similarity calculation method described above.

[0027] The output module is used to output the crystal structure whose similarity meets the preset conditions.

[0028] The fifth aspect of this application provides an electronic device, comprising:

[0029] Processor; and

[0030] A memory that stores executable code, which, when executed by the processor, causes the processor to perform the method described above.

[0031] A sixth aspect of this application provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method described above.

[0032] The technical solution provided in this application may include the following beneficial effects:

[0033] The method provided in this application obtains two crystal structure models, calculates the vector formed by element pairs of each matching type in each crystal structure model, and then calculates the target value density distribution. Based on the target value density distribution corresponding to the vector formed by element pairs of each matching type in each crystal structure model, the similarity between the two crystal structure models is determined. This method can calculate the similarity between two crystal structures, reducing computational resources and computation time.

[0034] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0035] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments thereof.

[0036] Figure 1 This is a schematic flowchart illustrating the crystal structure similarity calculation method shown in the embodiments of this application;

[0037] Figure 2 This is another schematic flowchart illustrating the crystal structure similarity calculation method shown in the embodiments of this application;

[0038] Figure 3 This is a schematic diagram of a copper crystal structure model shown in an embodiment of this application;

[0039] Figure 4 This is a schematic diagram illustrating the temperature change during the annealing simulation of a copper crystal structure model, as shown in an embodiment of this application.

[0040] Figure 5 This is a schematic diagram illustrating the density change during the annealing simulation of a copper crystal structure model, as shown in an embodiment of this application.

[0041] Figure 6This is a schematic diagram illustrating the changes in the similarity of vector length density distribution and vector angle density distribution during the annealing simulation of a copper crystal structure model, as shown in the embodiments of this application.

[0042] Figure 7 This is a schematic diagram of the system architecture at different simulation times during the annealing simulation of the copper crystal structure model, as shown in the embodiments of this application.

[0043] Figure 8 This is a schematic diagram of the vector length density distribution and vector angle density distribution at different simulation times during the annealing simulation of the copper crystal structure model shown in the embodiments of this application;

[0044] Figure 9 This is a schematic diagram of the XXIIIform C molecular crystal structure model shown in the embodiments of this application;

[0045] Figure 10 This is a schematic diagram illustrating the temperature change during the heating simulation of the XXIIIform C molecular crystal structure model, as shown in the embodiments of this application.

[0046] Figure 11 This is a schematic diagram illustrating the change in the similarity of vector length density distribution during the temperature simulation of the XXIIIform C molecular crystal structure model, as shown in the embodiments of this application.

[0047] Figure 12 This is a schematic diagram illustrating the change in the similarity of vector angle density distribution during the temperature simulation of the XXIIIform C molecular crystal structure model, as shown in the embodiments of this application.

[0048] Figure 13 This is a schematic diagram showing the comparison of molecular structures at different temperatures during the temperature simulation of the XXIIIform C molecular crystal structure model, as illustrated in the embodiments of this application.

[0049] Figure 14 This is a schematic diagram illustrating the change of lattice constant during the temperature simulation of the XXIIIform C molecular crystal structure model, as shown in the embodiments of this application.

[0050] Figure 15 This is a schematic diagram of the crystal structure similarity calculation device shown in the embodiments of this application;

[0051] Figure 16 This is another schematic diagram of the crystal structure similarity calculation device shown in the embodiments of this application;

[0052] Figure 17 This is a schematic flowchart illustrating the crystal structure screening method in an embodiment of this application;

[0053] Figure 18This is a schematic diagram of the crystal structure screening device shown in the embodiments of this application;

[0054] Figure 19 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation

[0055] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0056] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0057] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0058] In related technologies, the similarity between two crystal structures is calculated based on the RMSD output. However, calculating the RMSD is computationally intensive, time-consuming, and requires significant computational resources, making it difficult to apply widely. This is especially true for crystal structures with large sizes and a high number of atoms, where the computational load becomes even greater.

[0059] To address the aforementioned issues, this application provides a method for calculating crystal structure similarity, which can calculate the similarity between two crystal structures, reducing computational resources and computation time.

[0060] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0061] Figure 1 This is a schematic flowchart illustrating the crystal structure similarity calculation method shown in the embodiments of this application.

[0062] See Figure 1 The method includes:

[0063] Step S101: Obtain two crystal structure models.

[0064] The two crystal structure models can be pre-constructed based on the two crystal structures to be compared. A crystal structure can refer to a structure in which microscopic material units (such as atoms, ions, molecules, etc.) are arranged according to certain rules. A crystal structure can refer to a periodic or aperiodic crystal structure based on the local environment of atoms.

[0065] The two crystal structures can have the same types and / or proportions of elements.

[0066] Step S102: Calculate the vector formed by element pairs of each matching type in each crystal structure model; wherein, an element pair includes two matching elements.

[0067] In this context, an element can be one of the following: an atom, a group of atoms, or a molecule. That is to say, the two matching elements in an element pair can be atoms, groups of atoms, or molecules. For example, in an element pair, the two matching elements are element A and element B. Element A and element B can both be atoms, or element A and element B can both be groups of atoms, or element A and element B can both be molecules.

[0068] Element pairs can include atom pairs, group pairs, and molecule pairs. An element pair is an atom pair when the two matching elements are both atoms. An element pair is a group pair when the two matching elements are both groups. An element pair is a molecule pair when the two matching elements are both molecules.

[0069] The element pair matching methods include: in a crystal structure model, selecting an element as the central element, and matching the central element with any element within a preset cutoff radius to obtain element pairs. Specifically, the center point of the central element (such as the geometric center, centroid, or weighted geometric center) can be used as a reference to obtain the center points of other elements within a preset cutoff radius. These other elements, whose center points are within the preset cutoff radius, can be paired with the central element to obtain multiple element pairs. By changing the central element, each matching type of element pair in the crystal structure model can be obtained. In this way, matching different elements can yield atomic pairs, atomic group pairs, and molecular pairs.

[0070] In one implementation, calculating the vector formed by element pairs of each matching type in each crystal structure model may include:

[0071] Obtain the center point of each element in each matching type of element pair in each crystal structure model, and determine the vector formed by the element pair based on the center points of the two elements in the element pair.

[0072] The center point of an element can be its geometric center, centroid, or weighted geometric center. An element can be an atom, a group of atoms, or a molecule.

[0073] In this step, when the element pairs are atomic pairs, after determining that the element types and / or element ratios are the same in the two crystal structure models, the vector formed by each matching type of element pairs in each crystal structure model can be calculated.

[0074] When the element pairs are either atomic group pairs or molecular pairs, the vector formed by each matching type of element pairs in each crystal structure model can be directly calculated. In other words, when the element pairs are either atomic group pairs or molecular pairs, it is not necessary to compare the element types and / or element ratios in the two crystal structure models; the vector formed by each matching type of element pairs in each crystal structure model can be directly calculated.

[0075] Step S103: Calculate the target value density distribution based on the vector formed by the element pairs of each matching type.

[0076] The target value density distribution may include the vector length density distribution of each vector, and / or the vector angle density distribution of the vector angle between the corresponding two vectors for any two elements of the same matching type, and / or the vector dot product density distribution.

[0077] In this step, the target value can be calculated based on the vector formed by the element pairs of each matching type. The target value includes the vector length of each vector, and / or the vector angle and / or the dot product between the corresponding vectors of any two element pairs of the same matching type. Based on the target value, the target value density distribution is calculated.

[0078] Step S104: Determine the similarity between two crystal structure models based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model.

[0079] When the element pair is a group pair, the starting or ending point of the vector can be determined based on the geometric center, centroid, or weighted geometric center of the group, thus determining the vector formed by the group pair. Similarly, when the element pair is a molecule pair, the starting or ending point of the vector can be determined based on the geometric center, centroid, or weighted geometric center of the molecule, thus determining the vector formed by the molecule pair.

[0080] In this step, the similarity of the target value density distribution of each matching type of element pair between the two crystal structure models can be calculated based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model. The similarity between the two crystal structure models is then determined based on the similarity of the target value density distribution of element pairs of different matching types between the two crystal structure models.

[0081] As can be seen from this embodiment, the method provided in this application obtains two crystal structure models, calculates the vector formed by element pairs of each matching type in each crystal structure model, and then calculates the target value density distribution. Based on the target value density distribution corresponding to the vector formed by element pairs of each matching type in each crystal structure model, the similarity between the two crystal structure models is determined. This allows for the calculation of the similarity between two crystal structures, reducing computational resources and computation time.

[0082] Figure 2 This is another schematic flowchart of the crystal structure similarity calculation method according to an embodiment of this application. Figure 2 relatively Figure 1 The scheme of this application is described in more detail. It should be noted that... Figure 2 The embodiments are described using element pairs as atomic pairs, but element pairs are not limited to atomic pairs. Replacing atomic pairs with other element pairs (such as group pairs, molecular pairs, etc.) also applies. Figure 2 Example.

[0083] See Figure 2 The method includes:

[0084] Step S201: Based on the two crystal structures to be compared, construct two corresponding crystal structure models.

[0085] Crystal structure can refer to a structure in which microscopic elemental units (such as atoms, ions, molecules, etc.) are arranged according to certain rules. Crystal structure can refer to periodic or aperiodic crystal structures based on the local atomic environment. For example, periodic or aperiodic structures with larger dimensions and a larger number of atoms. In the embodiments of this application, periodic or aperiodic structures with larger dimensions and a larger number of atoms refer to crystal structures with a number of atoms exceeding a set value (e.g., 1000).

[0086] The constructed crystal structure model can be either a primitive cellular structure model or a supercellular structure model. The supercellular structure model can be obtained by expanding the primitive cellular structure model; that is, by expanding the primitive cellular structure model by a predetermined factor, the supercellular structure model can be obtained.

[0087] Step S202: Obtain two crystal structure models.

[0088] It should be noted that two crystal structure models can include two structural models of the same crystalline substance under different environments. For example, two structural models of copper crystal under different temperature environments. Another example is two structural models of XXIIIform C molecular crystal under different temperature environments.

[0089] Step S203: Compare the types and / or proportions of elements in the two crystal structure models.

[0090] In this step, the types of elements in the two crystal structure models are compared to see if they are the same, and the proportions of each element in the two crystal structure models are the same. Based on the analysis results, either step S204 or step S205 is selected.

[0091] Step S204: After determining that the two crystal structure models have different types of elements or different proportions, determine that the two crystal structure models belong to different structures.

[0092] It is understandable that if two crystal structures have different types of elements, different proportions of elements, or different types and proportions of elements, then the two crystal structures can be identified as different structures.

[0093] In this step, if it is determined that the types or proportions of elements in the two crystal structure models are different, the calculation result is output. The calculation result is used to indicate that the two crystal structure models belong to different structures. At this time, the similarity between the two crystal structure models will not be calculated.

[0094] It should be noted that, for element pairs that are atomic group pairs or molecular pairs, it is not necessary to compare the element types and / or element ratios in the two crystal structure models. That is, ignore the comparison of element types and / or element ratios in the two crystal structure models, and directly calculate the vector formed by each matching type of element pair in each crystal structure model.

[0095] Step S205: After determining that the types and / or proportions of elements in the two crystal structure models are the same, calculate the vector formed by each matching type of atom pairs in each crystal structure model; wherein, an atom pair includes two matching atoms.

[0096] It is understandable that even if two crystal structures contain the same types and proportions of elements, they are not necessarily the same substance, and their similarity is not 100%. For example, allotropes have the same types and proportions of elements but belong to different substances; for example, graphite and diamond. Another example is the existence of polymorphs in some molecular crystals. In polymorphic structures, the types and proportions of elements in the molecules are the same, but the arrangement of the molecules can be different, such as different crystal forms of remdesivir. Therefore, the above examples illustrate that even if two crystal structures contain the same types and proportions of elements, they can still belong to different substances.

[0097] In one implementation, the matching method for atomic pairs may include: selecting one atom as the central atom in a crystal structure model, and matching the central atom with any atom within a predetermined cutoff radius to obtain an atomic pair. The central atom is then replaced, and this process is repeated to obtain atomic pairs of each matching type in the crystal structure model.

[0098] In this context, the atoms in two atomic pairs with different matching types have different element types. Furthermore, even if the atoms in two atomic pairs have the same element type, the matching type will be different if the element type of the central atom is different. Here, "element type" refers to the atom type.

[0099] For example, a crystal structure model may include multiple 'a' atoms and multiple 'b' atoms, where 'a' and 'b' atoms are different elements. In this crystal structure model, one 'a' atom can be selected as the central atom. This selected 'a' atom can be paired with any atom within a predetermined cutoff radius to obtain an atomic pair. If there are multiple 'a' atoms and multiple 'b' atoms within the predetermined cutoff radius of the central atom (i.e., the selected 'a' atom), then the atomic pair can be obtained by pairing the central atom with another 'a' atom, i.e., atomic pair aa. Alternatively, the atomic pair can be obtained by pairing the central atom with a 'b' atom, i.e., atomic pair ab.

[0100] In this crystal structure model, a b atom can be selected as the central atom, and this selected b atom can be paired with any atom within a predetermined cutoff radius to obtain an atom pair. If there are multiple a atoms and multiple b atoms within the predetermined cutoff radius of the central atom (i.e., the selected b atom), then the atom pair can be obtained by pairing the central atom with another b atom, i.e., atom pair bb. Alternatively, the atom pair can be obtained by pairing the central atom with an atom, i.e., atom pair ba.

[0101] It is understandable that the central atom can be matched with any number of atoms within a predetermined cutoff radius to obtain multiple atom pairs. In a crystal structure model, all atoms can be selected sequentially as the central atom, and matched with any atom within the predetermined cutoff radius to obtain atom pairs.

[0102] In the example above, atom pair 'aa' is one type of matching, while atom pair 'ab' is another. Atom pair 'aa' and atom pair 'ab' are atomic pairs of different matching types. In other words, the atoms in two atomic pairs of different matching types have different element types.

[0103] Furthermore, in the matched atom pairs, the elemental types of both atoms fall within a predetermined element set. This predetermined element set is determined based on the elemental types in the two crystal structure models, and it includes at least some of the elements from those models. In other words, the elemental types in the predetermined element set can be all the elemental types in the two crystal structure models, or only some of them. For example, if both crystal structure models include three elemental types a, b, and c, then the predetermined element set can include a, b, and c; alternatively, it can include only a and b.

[0104] Furthermore, for atom pair ab, the central atom of atom pair ab is atom a. If atom b is taken as the central atom and matched with an atom within a predetermined cutoff radius, we obtain atom pair ba. Then atom pair ab and atom pair ba are atom pairs of different matching types. That is to say, when the atoms in two atom pairs have the same element type, if the element type of the central atom is different, the matching type is different.

[0105] It should be noted that when the element is an atom, the number of matching types is determined by the number of elemental types (i.e., the number of atom types) in the crystal structure model. In one implementation, when there are n types of atoms, the number of matching types can reach at most n squared (i.e., n... 2In this context, n is a positive integer. For example, when there are two types of atoms, the number of matching types can be four. For instance, assuming the crystal structure model includes two element types, a and b, the matching atom pairs can include: aa, ab, ba, and bb, which represent four matching types. As another example, assuming the crystal structure model includes three element types, a, b, and c, the matching atom pairs can include: aa, ab, ac, bb, ba, bc, cc, ca, and cb, which represent nine matching types. Similarly, when the element is a group of atoms, the number of matching types is determined by the number of group types in the crystal structure model; when the element is a molecule, the number of matching types is determined by the number of molecule types in the crystal structure model.

[0106] The value of the preset cutoff radius can be determined based on the analysis of the crystal structure of the crystal structure module, and the preset cutoff radius can also be set between 6 angstroms and 15 angstroms.

[0107] It is understandable that an atomic pair can determine a vector; for example, an atomic pair ab can determine a vector v. ab Vector v ab The direction can be from the central atom (atom a) to atom b. Vector v ab The length is the distance between the central atom (atom a) and atom b, i.e., the length l. ab .

[0108] Step S206: Calculate the target value based on the vector formed by the atomic pairs of each matching type.

[0109] The target value includes the vector length of each vector, and / or the vector angle and / or vector inner product between the corresponding vectors of any two atom pairs of the same matching type.

[0110] In one implementation, the target value is the vector length of each vector. For each type of matching atom pair, the vector length of each atom pair vector corresponding to that matching type is calculated.

[0111] For example, given the atom pair 'aa', there are 3 such matching atom pairs. Therefore, calculating the vector length of the vector formed by the 3 atom pairs 'aa' yields 3 different vector lengths 'l'. aa .

[0112] In another implementation, the target value is the vector angle between the two vectors corresponding to any two atomic pairs of the same matching type. For each type of matching type of atomic pairs, the vector angle between the two vectors corresponding to any two atomic pairs of that matching type is calculated.

[0113] For example, given an atom pair 'aa', there are three such pairs: the first atom pair 'aa', the second atom pair 'aa', and the third atom pair 'aa'. To calculate the angle between the corresponding vectors of the first and second atom pairs 'aa', the angle between the corresponding vectors of the second and third atom pairs 'aa', and the angle between the corresponding vectors of the first and third atom pairs 'aa', we obtain three different vector angles 'a'. aa .

[0114] In another implementation, the target value is the vector dot product between the two vectors corresponding to any two atomic pairs of the same matching type. For each type of matching type of atomic pairs, the vector dot product between the two vectors corresponding to any two atomic pairs of that matching type is calculated.

[0115] In another implementation, the target value includes the vector length of each vector, the vector angle between the two vectors corresponding to any two atom pairs of the same matching type, and the vector inner product between the two vectors corresponding to any two atom pairs of the same matching type.

[0116] In another implementation, the target values ​​include the vector lengths of each vector and the vector angle between the two vectors corresponding to any two atomic pairs of the same matching type. For each type of matching type of vector formed by atomic pairs, the vector lengths of the vectors of each atomic pair of the corresponding matching type are calculated, and the vector angle between the two vectors corresponding to any two atomic pairs of the corresponding matching type is calculated.

[0117] For example, given an atom pair 'aa', there are three such pairs: the first, second, and third atom pairs 'aa'. To calculate the length of the vector formed by these three pairs, we need to calculate the vector angles between corresponding vectors from the first and second atom pairs 'aa', the vector angles between corresponding vectors from the second and third atom pairs 'aa', and so on. This will yield three different vector lengths 'l'. aa And the angle α between three different vectors. aa .

[0118] Step S207: Calculate the target value density distribution based on the target value.

[0119] Among them, the target value density distribution is the density distribution of the target value, or the probability density distribution of the target value.

[0120] In this step, based on the target value, a preset density distribution calculation algorithm (such as the Gaussian kernel density estimation algorithm) can be used to calculate the density distribution of the target value. In one embodiment, if the target value is a vector length l, then the density distribution d of the vector length is calculated. l In another implementation, if the target value is the vector angle 'a', then the density distribution 'd' of the vector angle is calculated. a In another implementation, if the target value includes the vector length l and the vector angle α, then the density distribution d of the vector length is calculated. l and the density distribution d of the vector angle a .

[0121] It can be understood that the target value density distribution corresponds to the target value calculated in step S206. The target value density distribution can be a vector length density distribution, a vector angle density distribution, or a combination of both.

[0122] Step S208: Based on the target value density distribution corresponding to the vector formed by the atomic pairs of each matching type in each crystal structure model, calculate the similarity of the target value density distribution of the atomic pairs of each matching type between the two crystal structure models.

[0123] It is understandable that a vector formed by a pair of atoms of a certain matching type can determine the target value density distribution in two crystal structure models. For example, for a pair of atoms ab, the vector v formed by the pair of atoms of this matching type... ab It can determine the target value density distribution (i.e., vector length density distribution, and / or vector angle density distribution, and / or vector inner product density distribution) in one crystal structure model, and it can also determine the target value density distribution in another crystal structure model.

[0124] Based on the target value density distribution corresponding to the vectors formed by a matching type of atom pairs in two crystal structure models, the similarity of the target value density distribution of the matching type of atom pairs between the two crystal structure models can be calculated.

[0125] For example, two crystal structure models are crystal structure model A and crystal structure model A′, where, for atom pair ab, the vector v formed by this matching type of atom pair in crystal structure model A.ab The vector v that corresponds to the target value density distribution d in the A′ crystal structure model ab The corresponding target value density distribution d′. Based on the target value density distribution d corresponding to crystal structure model A and the target value density distribution d′ corresponding to crystal structure model A′, the similarity of the target value density distributions of element a and element b in atom pair ab between the two crystal structure models (i.e., crystal structure model A and crystal structure model A′) can be calculated.

[0126] It can be understood that a single matching type of atomic pair can correspond to a single target value density distribution similarity. Multiple matching types of atomic pairs can correspond to multiple target value density distribution similarities. For example, if two crystal structure models have atomic pairs of the matching type "aa", then the two crystal structure models can determine a single target value density distribution similarity. As another example, if two crystal structure models have atomic pairs of four different matching types—aa, ab, ba, and bb—the two crystal structure models can determine four different target value density distribution similarities.

[0127] The similarity of the target value density distribution corresponds to the target value density distribution itself. If the target value density distribution is a vector length density distribution, then the similarity corresponds to the vector length density distribution. If the target value density distribution is a vector angle density distribution, then the similarity corresponds to the vector angle density distribution. If the target value density distribution is a vector dot product density distribution, then the similarity corresponds to the vector dot product density distribution. If the target value density distribution includes both vector length density distribution and vector angle density distribution, then the similarity corresponds to the vector length density distribution and the vector angle density distribution, respectively. If the target value density distribution includes vector length density distribution, vector angle density distribution, and vector dot product density distribution, then the similarity corresponds to the vector length density distribution, the vector angle density distribution, and the vector dot product density distribution, respectively. It should be noted that different preset similarity calculation algorithms can be used to calculate the similarity of different types of density distributions. In one implementation, the F-divergence algorithm is used to calculate the similarity of the vector length density distribution and the vector angle density distribution.

[0128] Step S209: Determine the similarity between the two crystal structure models based on the similarity of the target value density distribution of atomic pairs of different matching types between the two crystal structure models.

[0129] In one implementation, this step may include:

[0130] Step S209-1: Calculate the weighted average of the similarity of the target value density distribution between the two crystal structure models based on the similarity of the target value density distribution of atomic pairs of different matching types.

[0131] It is understandable that in the process of calculating the weighted average, when all weight coefficients are the same, the weighted average is the average value.

[0132] For example, if two crystal structure models have four matching types of atomic pairs: aa, ab, ba, and bb, and the two crystal structure models determine four target value density distribution similarities, then these four target value density distribution similarities can be averaged to obtain the average target value density distribution similarity. Furthermore, the weighting coefficients of the four target value density distribution similarities can be adjusted as needed to perform a weighted average, obtaining the weighted average target value density distribution similarity. Further, when the target value density distribution similarity includes vector length density distribution similarity and vector angle density distribution similarity, the weighted average of the vector length density distribution similarity and vector angle density distribution similarity can be calculated as the output of the weighted average target value density distribution similarity calculation.

[0133] For example, if two crystal structure models have atomic pairs of the matching type aa, and the two crystal structure models determine a similarity of vector length density distribution and a similarity of vector angle density distribution, then the similarity of vector length density distribution and the similarity of vector angle density distribution can be weighted and averaged to calculate the weighted average of the similarity of the target value density distribution.

[0134] Furthermore, a weighted average of the similarity of the target value density distribution can be calculated based on the similarity of the target value density distribution between two crystal structure models for a preset number of atomic pairs of different matching types.

[0135] For example, if two crystal structure models have four matching types of atomic pairs: aa, ab, ba, and bb, and the two crystal structure models determine four target value density distribution similarities, then the similarities of three of the target value density distributions can be weighted and averaged to obtain the weighted average of the target value density distribution similarities; alternatively, the similarities of two of the target value density distributions can be weighted and averaged to obtain the weighted average of the target value density distribution similarities.

[0136] Step S209-2: Determine the similarity between the two crystal structure models based on the weighted average of the similarity of the target value density distribution.

[0137] In one embodiment of this step, the weighted average of the similarity of the target value density distribution can be output as the calculation result of the similarity between the two crystal structure models.

[0138] In other embodiments, in step S209, the similarity of different target value density distributions can be comprehensively evaluated based on the similarity of the target value density distributions of atomic pairs of different matching types between the two crystal structure models, and the similarity of the two crystal structure models can be determined. That is, instead of performing a weighted average processing on the various different target value density distribution similarities, the various different target value density distribution similarities are directly output as the calculation result of the similarity between the two crystal structure models.

[0139] As can be seen from this embodiment, the method provided in this application can calculate the similarity between two crystal structures, thereby reducing computational resources and computation time.

[0140] To illustrate this application as follows Figure 1 or Figure 2 The method provided in the embodiments can effectively and reliably calculate the similarity between two crystal structures. The structural changes of copper crystals during the annealing process are used as an example for detailed explanation.

[0141] In this embodiment, the structural changes of copper crystals during the annealing process are simulated using molecular dynamics simulation software.

[0142] It should be noted that copper crystals have a face-centered cubic structure with a cell edge length of 3.6147 angstroms. Please refer to [link / reference]. Figure 3 , Figure 3 In the diagram, 'a' represents the unit cell structure model of a copper crystal. Figure 3 In Figure b, the copper crystal supercell structure model is shown. This copper crystal supercell structure is a 10*10*10 scale structure, obtained by expanding the copper crystal primitive cell structure model. This copper crystal supercell structure model contains 4000 copper atoms.

[0143] In one implementation, based on a pre-built, such as Figure 3 The copper crystal supercell structure model shown in Figure b uses the EAM force field to describe the interaction between copper atoms and employs LAMMPS molecular dynamics simulation software for molecular dynamics simulation.

[0144] The copper crystal annealing simulation process is as follows: First, energy minimization calculations are performed on the initial copper crystal supercell structure model, and a stable copper crystal supercell structure model is obtained after optimization. Then, the model is subjected to continuous heating simulation, from 200K to 1500K, with a simulation heating step count of 8,000,000 steps. Next, it isothermal relaxation is performed at 1500K for 1,000,000 steps. After relaxation, continuous cooling simulation is performed, from 1500K to 200K, with a simulation step count of 8,000,000 steps. The time step in the simulation is 1 fs (i.e., each step lasts 1 fs), and the ensemble implementation method is NPT (isobaric isothermal).

[0145] Based on trajectory structure analysis and calculations of the copper crystal supercell model, the cutoff radius of the interatomic vector length in the copper crystal supercell model is 6 angstroms. This can be achieved by executing the application... Figure 1 or Figure 2 The method provided in the embodiment calculates the vector length density distribution and the vector angle density distribution, and uses the F-divergence calculation algorithm to calculate the similarity between the vector length density distribution and the vector angle density distribution, using the copper crystal supercell structure model at the moment when the temperature first reaches 300K as a comparison reference.

[0146] Relevant data from the simulation results of the copper crystal annealing process are as follows: Figures 4 to 8 As shown. Among them, Figure 4 The temperature variation of the copper crystal supercell structure model system with simulation time is shown. Figure 5 The process of density variation of the copper crystal supercell structure model system with simulation time is shown. Figure 6 The simulation demonstrates the changes in the similarity between the vector length density distribution and the vector angle density distribution of the copper crystal supercell structure model at the moment it first reaches a temperature of 300K, using the model as a comparative reference. Figure 7 The system architecture of the copper crystal supercell model is shown at different simulation times. Figure 8 Figure a shows the vector length density distribution of copper atom pairs in the architecture of the copper crystal supercell model at different simulation times. Figure 8 Figure b shows the vector angle density distribution of copper atom pairs in the architecture of the copper crystal supercell model at different simulation times.

[0147] It can be observed that in the first simulation phase (corresponding to times 0ns to 7.5ns), such as Figures 4 to 6As shown, after the simulation started, with the increase of temperature, the architecture of the copper crystal supercell model gradually changed, the density of the architecture gradually decreased, and the structure began to deviate from the structure at room temperature. At approximately 7.5 ns and a temperature of approximately 1300 K, the density of the architecture experienced a jump, and the similarity also jumped. This jump corresponds to the melting of the copper crystal supercell model during the heating process. According to... Figure 8 The vector length density distribution and vector angle density distribution of copper atom pairs in the copper crystal supercell structure model at different simulation times show that when the copper crystal structure transitions to the molten state (around 7.5 ns), both the length and angle density distributions undergo significant changes. The curves in the figure change from sharp peaks to smoother peaks or even disappear. Combined with... Figure 6 The changes in the similarity parameters shown reflect the transformation from a regularly arranged atomic structure to a disordered one in the copper crystal structure. (See also...) Figure 7 It is evident that the supercell structure model of copper crystals maintains a face-centered cubic trajectory structure at 7 ns, while the atomic arrangement within the crystal structure becomes more random at 9 ns. Figure 6 and Figure 8 The changes in density distribution and similarity shown correspond to this process.

[0148] In the second simulation phase (corresponding to times 7.5 ns to 13 ns), the density of the copper crystal structure system continuously increases as the temperature decreases. However, the density distribution of the vector length and vector angle of the system structure does not change significantly, and the similarity parameter also does not change significantly, remaining consistent with the size of the molten copper structure at high temperature. Figure 7 The copper crystal supercell model shown exhibits atoms in a disordered state at 9 ns, 11 ns, and 13 ns, indicating that the copper crystal system remains in a molten state despite the decreasing temperature between 8 ns and 13 ns. This demonstrates that... Figure 6 and Figure 8 The changes in density distribution and similarity shown correspond to this process.

[0149] In the third simulation phase (corresponding to times 13ns to 17ns), the copper crystal supercell model exhibits a sudden increase in system density at approximately 14ns and 750K after a further decrease in temperature. This change indicates a shift in the system architecture at this point. Figure 6 The magnitude of the structural similarity shown indicates that the copper crystal supercell structure model has recrystallized into a crystalline state, and the crystal form remains consistent with that of the unmelted state. (Comparison) Figure 7The morphology of the copper crystal at 15 ns and 17 ns can be determined; the structure is a regularly arranged atomic crystal structure, and the crystal form is face-centered cubic. Therefore, Figure 6 and Figure 8 The changes in density distribution and similarity shown correspond to this process.

[0150] In summary, the crystal structure similarity calculation method provided in this application can effectively and reliably calculate the similarity between two crystal structures. When applied to the study of the structural changes of copper crystals during the annealing process, the calculated similarity between the two crystal structures (the similarity between the copper crystal structure at different times during the temperature change process and the copper crystal structure at the moment when the temperature first reaches 300K) can objectively and accurately reflect the structural change process of copper crystals during the annealing process.

[0151] To further illustrate this application, Figure 1 or Figure 2 The method provided in the examples can effectively and reliably calculate the similarity between two crystal structures. The structural changes of the XXIIIform C molecular crystal (the XXXIIform C molecular crystal structure has a CDC number of 1447524 in the CCDC database) during the heating process are used as an example for detailed explanation.

[0152] In this embodiment, the structural changes of XXIIIform C molecular crystal during the heating process are simulated using molecular dynamics simulation software.

[0153] like Figure 9 As shown in Figure A, a single XXIIIform C molecular crystal contains 43 atoms, including five elements: O, N, C, H, and Cl. The space group of the XXIIIform C molecular crystal is p1, and the edge lengths of the primitive unit cell are 7.491 Å, 11.767 Å, ​​and 20.379 Å, respectively. The lattice angles are 87.15 degrees, 93.57 degrees, and 100.37 degrees, respectively, and the primitive unit cell contains 4 molecules. Figure 9 B in the model represents the unit cell structure of the XXIIIformC molecular crystal. Figure 9 C represents the supercell structure model of the XXIIIform C molecular crystal. This supercell structure is a 4*2*1 scale structure, obtained by expanding the original XXIIIform C molecular crystal structure model. The supercell structure model contains 32 molecules.

[0154] In one implementation, based on a pre-built, such as Figure 9The XXIIIform C molecular crystal supercell structure model shown in Figure C uses the Gaff force field to describe the intermolecular interactions and employs Gromacs molecular dynamics simulation software for molecular dynamics simulation.

[0155] The temperature-increasing simulation process for XXIIIform C molecular crystals is as follows: First, energy minimization calculations are performed on the initial supercell structure model of the XXIIIform C molecular crystal. After optimization, a stable supercell structure model of the XXIIIform C molecular crystal is obtained. Then, the model is continuously heated from 50K to 300K in 5,000,000 steps. The time step during the simulation is 1 fs, and the ensemble implementation method is NPT.

[0156] Based on the trajectory structure analysis of the four elements (O, N, C, and CI, excluding H) in the XXIIIform C molecular crystal supercell structure model, the cutoff radius of the interatomic vector length in the XXIIIform C molecular crystal supercell structure model is 10 Å. It should be noted that the atomic mass of H is relatively small, and its displacement variation during molecular dynamics simulations may be large, making the structural similarity calculated from the vectors between H and other atoms difficult to reflect the actual structural similarity. Therefore, the contribution of H atoms is generally not considered when calculating structural similarity. Furthermore, the position of H atoms is difficult to observe in commonly used X-ray diffraction methods for determining structure; therefore, H atoms are removed. This application... Figure 1 or Figure 2 The method provided in the embodiment calculates the vector length density distribution and the vector angle density distribution, and uses the F-divergence calculation algorithm to calculate the similarity of the vector length density distribution and the vector angle density distribution by using the XXIIIform C molecular crystal supercell structure model at the moment when the temperature first reaches 200K as a comparison reference.

[0157] The relevant data from the simulation results of the annealing process of XXIIIform C molecular crystal are as follows: Figures 10 to 14 As shown. Among them, Figure 10 The temperature variation of the XXIIIform C molecular crystal supercell structure model system with simulation time is shown. Figure 11 The study demonstrates the variation of the similarity of vector length density distributions corresponding to vectors formed by different atomic pairs in the XXIIIform C molecular crystal supercell structure model with temperature. Figure 12 The similarity of the vector angle density distribution corresponding to the vectors formed by different atomic pairs in the XXIIIform C molecular crystal supercell structure model is shown as a function of temperature. Figure 13 Image A shows a comparison of the molecular structures in the supercell models of XXIIIform C molecular crystals at 50K and 200K. Figure 13 Figure B shows a comparison of the molecular structure in the supercell structure model of the XXIIIform C molecular crystal at 200K and 300K temperatures. Figure 14 The variation of the lattice constant with temperature in the supercell structure model of the XXIIIform C molecular crystal is shown.

[0158] It can be observed that the structure of XXIIIform C molecular crystals is relatively stable in the low-temperature range of 50K-100K, without significant changes. Within this temperature range, the similarity in the density distribution of the lengths and angles of the vectors formed by different atomic pairs within the structure does not change significantly. At approximately 110K, combined with… Figure 11 and Figure 12 It can be seen that the similarity of the XXIIIform C molecular crystal structure changed abruptly, indicating that the crystal structure underwent certain changes at higher temperatures. The similarity in the density distribution of the lengths and angles of the vectors formed by the atomic pairs shows that it is mainly the chlorine-oxygen atomic pairs (…). Figure 11 and Figure 12 (The curve indicated by ① in the middle) and chlorine-nitrogen atom pairs ( Figure 11 and Figure 12 The density distribution similarity of the curves indicated by marker ② has changed significantly. Figure 11 and Figure 12 The curve indicated by symbol ③ reflects the overall change in the similarity of the density distribution of the lengths and angles of the vectors corresponding to all atomic pairs with temperature. According to... Figure 13 A comparison of the molecular structures in the XXIIIform C molecular crystal supercell models at 50K and 200K, shown in Figure A, reveals that the position of chlorine in the molecular structure shifts at 200K compared to 50K. This shift is primarily due to changes in the dihedral angle between the benzene ring connected to chlorine and the adjacent benzene ring. Consequently, the relative positions of chlorine and the surrounding oxygen and nitrogen atoms change, leading to alterations in the length and angle of the vectors formed by chlorine-oxygen and chlorine-nitrogen atom pairs.

[0159] As the temperature continued to rise, at approximately 275 K, the similarity between the crystal structure and the crystal structure at 200 K changed dramatically. Figure 11 and Figure 12 It can be seen that the density distribution similarity of chlorine-oxygen atom pairs and chlorine-nitrogen atom pairs still has the greatest impact on the overall density distribution similarity of all atom pairs. According to Figure 13Comparing the molecular structures in the XXIIIform C supercell models at 200K and 300K as shown in Figure B, it can be seen that the molecular structure at 300K, compared to that at 200K, exhibits a significant structural change, with the dihedral angle between the benzene ring connected to chlorine and the adjacent benzene ring shifting by approximately 90 degrees. The positions of the chlorine atoms also change considerably compared to the 200K molecular structure. This leads to substantial alterations in the length and angle of the vectors formed by the chlorine-oxygen and chlorine-nitrogen atom pairs between adjacent molecules, resulting in a jump in structural similarity at 275K.

[0160] according to Figure 14 The variation of the lattice constant with temperature in the supercell structure model of the XXIIIform C molecular crystal shows that at approximately 110 K, the lattice constant does not change significantly; therefore, the overall crystal structure is merely a normal change under the influence of increasing temperature. However, at 275 K, the molecular structure undergoes a significant change, and the cell constant also changes considerably. Therefore, it can be determined that at 275 K, the XXIIIform C molecular crystal undergoes a transformation. This demonstrates that… Figure 11 and Figure 12 The change process of the density distribution similarity shown corresponds to this.

[0161] In summary, the crystal structure similarity calculation method provided in this application can effectively and reliably calculate the similarity between two crystal structures. When applied to the study of the structural changes of XXIIIform C molecular crystals during the heating process, the calculated similarity between the two crystal structures (the similarity between the XXIIIform C molecular crystal structure at different times during the temperature change process and the XXIIIform C molecular crystal structure at the moment when the temperature is first reached 200K) can objectively and accurately reflect the structural change process of XXIIIform C molecular crystals during the heating process.

[0162] It should also be noted that, since the computational resources required for similarity calculation using RMSD in related technologies are proportional to the square of the number of atoms in the crystal structure, the computational resources required by the crystal structure similarity calculation method provided in this application are only proportional to the number of atoms in the crystal structure. This makes the crystal structure similarity calculation method provided in this application particularly suitable for calculating the similarity between two crystal structures with large sizes and a large number of atoms. Furthermore, when comparing the similarity between multiple crystal structures, the RMSD similarity calculation scheme in related technologies requires calculation for each pair of crystal structures to be compared, while the crystal structure similarity calculation method provided in this application only needs to first calculate the density distribution of the length and included angle of the vectors formed by the atomic pairs for each crystal structure, and then perform the similarity calculation between each pair of structures, thus greatly reducing the computational load. This reduces computational resources, decreases the computational load, and shortens the computation time.

[0163] Corresponding to the aforementioned application function implementation method embodiments, this application also provides a crystal structure similarity calculation device, electronic device, and corresponding embodiments.

[0164] Figure 15 This is a schematic diagram of the crystal structure similarity calculation device shown in the embodiments of this application.

[0165] See Figure 15 A crystal structure similarity calculation device 150 includes: an acquisition module 1510, a first calculation module 1520, a second calculation module 1530, and an output module 1540.

[0166] Module 1510 is used to acquire two crystal structure models.

[0167] The first calculation module 1520 is used to calculate the vector formed by each matching element pair in the crystal structure model acquired by each acquisition module 1510; wherein, the element pair includes two matching elements.

[0168] The second calculation module 1530 is used to calculate the target value density distribution based on the vector formed by the element pairs of each matching type calculated by the first calculation module 1520.

[0169] The target value density distribution includes the vector length density distribution of each vector, and / or the vector angle density distribution of the vector angle between any two elements of the same matching type and the vector dot product density distribution.

[0170] The output module 1540 is used to determine the similarity between two crystal structure models based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model calculated by the second calculation module 1530.

[0171] As can be seen from this embodiment, the device 150 provided in this application embodiment can calculate the similarity between two crystal structures, reduce computing resources, and reduce computing time.

[0172] Figure 16 This is another schematic diagram of the crystal structure similarity calculation device shown in the embodiments of this application;

[0173] See Figure 16 A crystal structure similarity calculation device 150 includes: an acquisition module 1510, a first calculation module 1520, a second calculation module 1530, an output module 1540, a construction module 1550, and an analysis module 1560.

[0174] The functions of the acquisition module 1510, the first calculation module 1520, the second calculation module 1530, and the output module 1540 can be found in [reference needed]. Figure 15 The description in the text will not be repeated here.

[0175] Module 1550 is used to construct two crystal structure models based on the two crystal structures to be compared.

[0176] Analysis module 1560 is used to compare the types and / or proportions of elements in two crystal structure models.

[0177] Furthermore, the first calculation module 1520 is also used to calculate the vector formed by each matching type of element pair in each crystal structure model after determining that the element types and / or element ratios in the two crystal structure models are the same.

[0178] The first calculation module 1520 is also used to obtain the center points of the elements contained in each matching type of element pair in each crystal structure model. The vector formed by the element pair is determined based on the center points of the two elements in the element pair. The center points of the elements include the geometric center, centroid, or weighted geometric center of the element.

[0179] The first calculation module 1520 is also used to select an atom as the central atom in a crystal structure model, and match the central atom with any atom within a preset cutoff radius to obtain an atom pair. By changing the central atom, the atom pairs of each matching type in the crystal structure model are obtained.

[0180] The number of matching types is determined by the elemental composition of the atoms in the crystal structure model. The atoms in two atom pairs with different matching types have different elemental compositions. Furthermore, even when the atoms in two atom pairs have the same elemental composition, if the central atom has a different elemental composition, the matching types will be different.

[0181] In this case, the two atoms in an atom pair can belong to a preset set of elements.

[0182] The second calculation module 1530 is further configured to calculate a target value based on the vector formed by the element pairs of each matching type. The target value includes the vector length of each vector, and / or the vector angle and / or the dot product between the corresponding vectors of any two element pairs of the same matching type. Based on the target value, the target value density distribution is calculated.

[0183] The output module 1540 is also used to calculate the similarity of the target value density distribution of each matching type of element pair between two crystal structure models based on the target value density distribution corresponding to the vector formed by the element pairs of each matching type in each crystal structure model. The similarity between the two crystal structure models is determined based on the similarity of the target value density distribution of element pairs of different matching types between the two crystal structure models.

[0184] The output module 1540 is also used to calculate a weighted average of the similarity of the target value density distribution between two crystal structure models based on the similarity of the target value density distribution between element pairs of different matching types. The similarity between the two crystal structure models is then determined based on this weighted average of the target value density distribution similarity.

[0185] The output module 1540 is also used to determine that two crystal structure models belong to different structures after it is found that the types of elements in the two crystal structure models are different or the proportions of elements are different.

[0186] Regarding the apparatus 150 in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0187] Figure 17 This is a schematic flowchart illustrating the crystal structure screening method in an embodiment of this application.

[0188] See Figure 17 The method includes:

[0189] Step S1701: Obtain multiple crystal structures.

[0190] Step S1702: Calculate the similarity between any two crystal structures using the crystal structure similarity calculation method.

[0191] For methods of calculating crystal structure similarity, please refer to [examples omitted]. Figure 1 or Figure 2 The relevant descriptions in the embodiments.

[0192] Step S1703: Output the crystal structure whose similarity meets the preset conditions.

[0193] In this step, different preset conditions can be set to output different results. For example, it can output the two crystal structures with the highest similarity; it can also output multiple sets of crystal structures with a preset number or preset proportion of similarity, arranged from largest to smallest; and it can also output the two crystal structures with the lowest similarity.

[0194] As can be seen from this embodiment, the method provided in this application makes it easier to screen out the target crystal structure from multiple crystal structures, thereby improving the screening efficiency and reliability of crystal structures.

[0195] Figure 18 This is a schematic diagram of the crystal structure screening device shown in the embodiments of this application.

[0196] See Figure 18 A crystal structure screening device 1800 includes: an acquisition module 1810, a calculation module 1820, and an output module 1830.

[0197] The acquisition module 1810 is used to acquire multiple crystal structures.

[0198] The calculation module 1820 is used to calculate the similarity between any two crystal structures using crystal structure similarity calculation methods.

[0199] Output module 1830 is used to output crystal structures whose similarity meets preset conditions.

[0200] As can be seen from this embodiment, the device 1800 provided in this application embodiment can screen out the target crystal structure from multiple crystal structures, thereby improving the crystal structure screening efficiency and screening reliability.

[0201] Figure 19 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application.

[0202] See Figure 19 The electronic device 1900 includes a memory 1910 and a processor 1920.

[0203] The processor 1920 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0204] Memory 1910 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by processor 1920 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 1910 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory 1910 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital versatile optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, an ultra-high density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0205] The memory 1910 stores executable code, which, when processed by the processor 1920, can cause the processor 1920 to execute some or all of the methods described above.

[0206] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0207] Alternatively, this application may be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of an electronic device (or server, etc.), causes the processor to perform part or all of the steps of the methods described above according to this application.

[0208] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method of calculating a degree of similarity of crystal structures, characterized by, The method comprises the following steps: obtaining two crystal structure models; calculating vectors formed by each matching type of element pair in each of the crystal structure models; wherein, obtaining the center point of each element included in each matching type of element pair in each of the crystal structure models; determining the vector formed by the element pair according to the center points of the two elements in the element pair; the element pair comprises two matching elements, and the element comprises an atom, an atomic group or a molecule; calculating a target value density distribution according to the vectors formed by each matching type of element pair; wherein, calculating a target value according to the vectors formed by each matching type of element pair; the target value comprises the vector length of each vector, and / or the vector included angle and / or vector inner product between any two vectors corresponding to two element pairs of the same matching type; calculating a target value density distribution according to the target value; determining the similarity of the two crystal structure models according to the target value density distribution corresponding to the vectors formed by each matching type of element pair in each of the crystal structure models.

2. The method of claim 1, wherein, The center point of the element comprises the geometric center, the center of mass or the weighted geometric center of the element.

3. The method of claim 2, wherein, When the element is an atom, the element pair is an atomic pair, and the matching mode of the atomic pair comprises: in one of the crystal structure models, selecting an atom as a center atom, matching the center atom with any atom within a preset cutoff radius to obtain the atomic pair; replacing the center atom to obtain each matching type of atomic pair in the crystal structure model; wherein, the number of matching types is determined by the element type of the atom in the crystal structure model.

4. The method of claim 3, wherein: the element types of the two atoms in the atomic pair belong to a preset element set range.

5. The method according to any one of claims 1 to 4, characterized in that, Before the step of calculating the vector formed by each matching type of element pair in each of the crystal structure models, the method further comprises: comparing the element types and / or element proportions in the two crystal structure models; after determining that the element types and / or element proportions in the two crystal structure models are the same, performing the step of calculating the vector formed by each matching type of element pair in each of the crystal structure models.

6. The method according to any one of claims 1 to 4, characterized in that, The step of determining the similarity of the two crystal structure models according to the target value density distribution corresponding to the vectors formed by each matching type of element pair in each of the crystal structure models comprises: calculating the target value density distribution similarity of each matching type of element pair between the two crystal structure models according to the target value density distribution corresponding to the vectors formed by each matching type of element pair in each of the crystal structure models; determining the similarity of the two crystal structure models according to the target value density distribution similarity of different matching types of element pairs between the two crystal structure models.

7. The method of claim 6, wherein, The step of determining the similarity of the two crystal structure models according to the target value density distribution similarity of different matching types of element pairs between the two crystal structure models comprises: According to the target value density distribution similarity of the element pairs of different matching types between two of the crystal structure models, a target value density distribution similarity weighted average value is calculated; According to the target value density distribution similarity weighted average value, the similarity of the two of the crystal structure models is determined.

8. A method of screening for a crystal structure, characterized by, The method comprises the following steps: Obtaining a plurality of crystal structures; Using the crystal structure similarity calculation method of any one of claims 1-7 to calculate the similarity of any two of the crystal structures; Outputting the crystal structure whose similarity meets a preset condition.

9. A crystal structure similarity calculation apparatus characterized by comprising: The method comprises the following steps: An obtaining module is configured to obtain two crystal structure models; A first calculating module is configured to calculate a vector formed by an element pair of each matching type in each of the crystal structure models obtained by the obtaining module; wherein, a center point of an element included in each of the element pairs of each matching type in each of the crystal structure models is obtained; a vector formed by the element pair is determined according to the center points of the two elements in the element pair; the element pair comprises two matching elements, and the element comprises an atom, an atom group or a molecule; A second calculating module is configured to calculate a target value density distribution according to the vectors formed by the element pairs of each matching type calculated by the first calculating module; wherein, a target value is calculated according to the vectors formed by the element pairs of each matching type; the target value comprises a vector length of each of the vectors, and / or a vector included angle and / or a vector inner product between two of the vectors corresponding to any two of the element pairs of the same matching type; a target value density distribution is calculated according to the target value; An output module is configured to determine the similarity of the two of the crystal structure models according to the target value density distribution corresponding to the vector formed by the element pair of each matching type in each of the crystal structure models calculated by the second calculating module.

10. A crystal structure screening apparatus, characterized by, The method comprises the following steps: An obtaining module is configured to obtain a plurality of crystal structures; A calculating module is configured to use the crystal structure similarity calculation method of any one of claims 1-7 to calculate the similarity of any two of the crystal structures; An output module is configured to output the crystal structure whose similarity meets a preset condition.

11. An electronic device, comprising: The method comprises the following steps: A processor; And A memory having executable code stored thereon, which, when executed by the processor, causes the processor to execute the method of any one of claims 1-8. 12.A computer readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to execute the method of any one of claims 1-8.

Images