Information processing method, information processing system, and program

Abstract

This information processing method executed by a computer includes: acquiring first structure information indicating a first structure composed of a plurality of atoms, and target element information indicating a target element that is a target for which the existence probability is represented in a unit cell corresponding to the first structure; generating, by using the first structure information and the target element information, first atomic distribution information indicating the distribution of the existence probability in the unit cell corresponding to the first structure regarding a target atom that is at least one atom represented by the target element; and outputting the first atomic distribution information.

Description

Information Processing Method, Information Processing System, and Program

[0001] The present disclosure relates to a technique for predicting the distribution of the probability of the existence of atoms.

[0002] In Patent Document 1, based on first information regarding a plurality of polyhedra and second information regarding a plurality of atoms arranged in each of the plurality of polyhedra, a third structure indicating a crystal structure that can be obtained when each of the plurality of atoms is arranged in a three-dimensional structure in which the plurality of polyhedra are arranged is generated. A technique is disclosed.

[0003] International Publication No. 2023 / 188733

[0004] However, in the technique described in Patent Document 1, no consideration has been given to the technique for predicting the distribution of the probability of the existence of atoms.

[0005] The present disclosure provides a technique capable of predicting the distribution of the probability of the existence of atoms.

[0006] An information processing method according to an aspect of the present disclosure includes acquiring first structure information indicating a first structure composed of a plurality of atoms and target element information indicating a target element that is a target representing the existence probability in a unit cell corresponding to the first structure, and using the first structure information and the target element information to generate first atom distribution information indicating the distribution of the existence probability of target atoms, which are one or more atoms represented by the target element, in the unit cell corresponding to the first structure, and outputting the first atom distribution information.

[0007] According to the present disclosure, the distribution of the probability of the existence of atoms can be predicted.

[0008] This is a block diagram showing the overall configuration of an information processing system. This is a flowchart showing an example of a process performed by the information processing system. This is a diagram showing an example of an information input screen. This is a diagram showing an example of first structure information. This is a diagram schematically illustrating the process of generating second structure information based on first structure information and target element information. This is a diagram schematically illustrating the process of generating atomic site information based on second structure information. This is a diagram schematically illustrating the process of calculating atomic density distribution based on target element information, second structure information, and target number of atoms information. This is a diagram schematically illustrating the process of generating first atomic distribution information based on atomic site information and atomic density distribution information. This is a diagram schematically illustrating the process of generating second atomic distribution information based on atomic site information, atomic density distribution information, and second structure information. This is a diagram showing an example of an output screen. This is a diagram showing another example of second atomic distribution information. This is a flowchart showing the process of predicting the third structure. This is a diagram schematically illustrating the process of generating third structure information based on target element information, second structure information, and first atomic distribution information. This is a diagram to explain an example of the process of predicting ion conduction paths. This is a diagram to explain another example of the process of predicting ion conduction paths. This is a diagram to explain the process of predicting battery characteristics.

[0009] (Background to one aspect of this disclosure) Conventionally, techniques for predicting the characteristics of a battery based on the battery materials that constitute the battery are known. For example, a technique is known that reproduces the crystal structure of the battery materials that constitute the battery and evaluates the characteristics of the battery based on the reproduced crystal structure. However, battery materials are sometimes composed of highly mobile atoms. For example, the battery material of a lithium-ion battery contains lithium atoms, which are highly mobile atoms. When reproducing the crystal structure of a battery material, the question arises as to where to place the highly mobile atoms.

[0010] One possible approach is to randomly arrange highly mobile atoms within the crystal structure. Specifically, multiple crystal structures can be generated by randomly arranging highly mobile atoms. From these multiple crystal structures, the one with the most stable energy can be extracted, and the battery's characteristics can be evaluated based on that structure.

[0011] However, the above method requires the processor to calculate the energy of multiple crystal structures and then determine the most stable crystal structure based on these calculated energies. This increases the computational cost of the processor. The above method generates a crystal structure that assumes highly mobile atoms are reliably positioned in their designated locations. It is difficult to accurately predict battery characteristics using this structure.

[0012] The inventors of this invention have conducted extensive research on techniques for predicting battery characteristics and have found that by predicting the probability distribution of atom presence (called atomic distribution) from the crystal structure, and using this atomic distribution to predict battery characteristics, the computational cost of the processor does not increase, and the accuracy of predicting battery characteristics improves.

[0013] This disclosure is based on such findings.

[0014] (1) An information processing method in one aspect of the present disclosure includes: acquiring first structure information indicating a first structure composed of a plurality of atoms; target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure; generating first atom distribution information indicating the distribution of the probability of existence of one or more target atoms represented by the target element in a unit cell corresponding to the first structure using the first structure information and the target element information; and outputting the first atom distribution information.

[0015] This configuration outputs first atomic distribution information, which shows the distribution of the probability of existence of the target atom. This first atomic distribution information can be used, for example, when predicting the properties of the first structure.

[0016] (2) The information processing method described in (1) above, wherein the first structure is composed of a plurality of atoms represented by two or more elements, the target element is one of the two or more elements, and further includes generating second structure information showing a second structure obtained by removing the target atom from the first structure, generating second atom distribution information showing the distribution of the probability of existence of the target atom in the second structure using the second structure information and the target element information, and outputting the second atom distribution information.

[0017] This configuration outputs second-atom distribution information, which shows the distribution of the probability of existence of the target atom in the second structure. This second-atom distribution information can be used, for example, when predicting the properties of the first structure.

[0018] (3) The information processing method described in (1) above, wherein the first structure is composed of a plurality of atoms represented by elements, the target element is an element arranged in the first structure, and further includes generating second structure information showing a second structure in which the target atoms are not arranged in the first structure, generating second atom distribution information showing the distribution of the probability of existence of the target atoms in the second structure using the second structure information and the target element information, and outputting the second atom distribution information.

[0019] This configuration outputs second-atom distribution information, which shows the distribution of the probability of existence of the target atom in the second structure. This second-atom distribution information can be used, for example, when predicting the properties of the first structure.

[0020] (4) The information processing method described in (2) above may further include generating atomic site information by predicting the positions of one or more atomic sites that are candidates for the positions in which the target atoms are placed, based on the arrangement of the remaining atoms left in the second structure.

[0021] According to this configuration, the positions of one or more candidate atomic sites for the target atom are predicted based on the arrangement of the remaining atoms in the second structure. Therefore, the positions of one or more atomic sites can be predicted more accurately compared to predicting the positions of one or more atomic sites without utilizing the arrangement of the remaining atoms.

[0022] (5) In the information processing method described in (4) above, generating the second atomic distribution information may include associating the probability of existence of the target atom with each of the one or more atomic sites in the second structure, based on the second structure information, the target element information, and the atomic site information.

[0023] This configuration allows for the generation of second atomic distribution information in which the probability of the target atom's presence is superimposed on each of the one or more atomic sites in the second structure. In other words, it is possible to generate second atomic distribution information that clearly indicates which atomic site contains the target atom and with what probability.

[0024] (6) In the information processing method described in (4) or (5) above, generating the second atomic distribution information may include calculating the probability of existence of the target atom for each of the one or more atomic sites using a probability distribution model.

[0025] In this configuration, the probability of finding a target atom at each of the one or more atomic sites is calculated using a probability distribution model. This makes it possible to calculate the probability of finding a target atom at each atomic site more accurately.

[0026] (7) In the information processing method described in any one of (2) to (6) above, outputting the second atomic distribution information includes outputting an image representing the second atomic distribution information, wherein the second atomic distribution information represented by the image may be at least one of the following: (a) information showing a structure in which the target atoms are arranged in the second structure and a value indicating the probability of existence of each of the arranged target atoms is superimposed on each of the arranged target atoms; (b) information showing a structure in which the target atoms are arranged in the second structure and a graph indicating the probability of existence of each of the arranged target atoms is superimposed on each of the arranged target atoms; (c) information showing a structure in which the target atoms are arranged in the second structure and each of the target atoms is colored and the probability of existence of each of the target atoms is represented by the intensity of the color or the transparency of the color.

[0027] According to this configuration, at least one of the following images representing the second atomic distribution information is output: (a) an image visualizing the probability of the existence of the target atom at each atomic site using numerical values, (b) an image visualizing the probability of the existence of the target atom at each atomic site using a graph, or (c) an image visualizing the probability of the existence of the target atom at each atomic site using the intensity of color or transparency. This makes it possible for users to intuitively understand the probability of the existence of the target atom at each atomic site.

[0028] (8) The information processing method described in any one of (2) to (7) above may further include generating third structure information, which is a third structure composed of a plurality of atoms represented by one or more elements, wherein the target atoms are arranged in a probability corresponding to the probability of existence of the target atoms indicated by the second atom distribution information, using the target element information, the second structure information, and the first atom distribution information, and outputting the third structure information.

[0029] This configuration allows for the generation of third structural information that shows a third structure in which the target atoms are arranged with a probability corresponding to the probability of existence of the target atoms indicated by the second atomic distribution information.

[0030] (9) The information processing method described in any one of (2) to (8) above may further include generating ion conduction path information indicating the ion conduction path of the target atom by performing conduction path prediction processing based on the second atomic distribution information, and outputting the ion conduction path information.

[0031] With this configuration, the ion conduction path is predicted based on the second atom distribution information. That is, the ion conduction path is predicted after considering the distribution of the probability of existence of the target atom. Therefore, it is possible to predict the ion conduction path more accurately compared to when the ion conduction path is predicted without using the second atom distribution information.

[0032] (10) The information processing method described in any one of (2) to (9) above may further include generating a plurality of structural models in which the number of target atoms differs from each other based on the second atomic distribution information, generating characteristic information that shows the properties of the first structure by performing a physical property simulation using the plurality of structural models as input, and outputting the characteristic information.

[0033] With this configuration, the properties of the first structure are predicted based on the second atom distribution information. That is, the properties of the first structure are predicted after considering the distribution of the probability of existence of the target atom. Therefore, compared to predicting the properties of the first structure without using the second atom distribution information, the properties of the first structure can be predicted more accurately.

[0034] This disclosure can be implemented not only as an information processing method that performs the characteristic processing described above, but also as an information processing system having a characteristic configuration corresponding to the characteristic processing performed by the information processing method. It can also be implemented as a computer program that causes a computer to execute the characteristic processing included in such an information processing method. Therefore, the same effects as the above information processing method can be achieved in the following other embodiments.

[0035] (11) An information processing system according to another aspect of the present disclosure is an information processing system including a processor, the processor performing the following actions: acquiring first structure information indicating a first structure composed of a plurality of atoms, and target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure; using the first structure information and the target element information, generating first atom distribution information indicating the distribution of the probability of existence of one or more target atoms represented by the target element in a unit cell corresponding to the first structure; and outputting the first atom distribution information.

[0036] (12) A program according to yet another aspect of the present disclosure causes a computer to acquire first structural information indicating a first structure composed of a plurality of atoms, and target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure, generate first atom distribution information indicating the distribution of the probability of existence in a unit cell corresponding to the first structure for one or more target atoms represented by the target element in a unit cell corresponding to the first structure, and output the first atom distribution information.

[0037] It goes without saying that the program related to this disclosure can be distributed via computer-readable non-temporary recording media such as CD-ROMs, or via communication networks such as the Internet.

[0038] The embodiments of this disclosure will be described below with reference to the drawings. Note that the embodiments described below are all specific examples of this disclosure. The numerical values, shapes, components, steps, and order of steps shown in the following embodiments are examples and are not intended to limit this disclosure. Components in the following embodiments that are not described in the independent claim representing the highest-level concept will be described as optional components. In all embodiments, the contents of each can be combined.

[0039] (Embodiment 1) FIG. 1 is a block diagram showing the overall configuration of an information processing system 1 according to Embodiment 1 of the present disclosure. The information processing system 1 is a system for predicting the characteristics of battery materials. Specifically, it is a system for predicting the distribution of the existence probability of predetermined atoms from the crystal structure of battery materials in order to predict characteristics such as the physical properties of batteries.

[0040] The information processing system 1 is composed of an information processing terminal such as a personal computer. However, this is just an example. The information processing system 1 may be composed of a server (for example, a cloud server) including one or more computers. In addition, the information processing system 1 may be composed of an edge device such as a laptop personal computer or a tablet-type terminal, and a server connected to the edge device via a communication network such as the Internet.

[0041] As shown in FIG. 1, the information processing system 1 includes an input unit 10, a processor 20, a storage unit 30, and a display unit 40.

[0042] The input unit 10 includes, for example, a mouse, a keyboard, and a touch panel.

[0043] The processor 20 includes, for example, a central processing unit (CPU) and an image processing unit (GPU).

[0044] The storage unit 30 includes a non-volatile rewritable storage device such as a hard disk drive or a solid state drive.

[0045] The display unit 40 includes a display such as a liquid crystal monitor.

[0046] The processor 20 includes an acquisition unit 21, a processing unit 22, and an output unit 23. The acquisition unit 21 to the output unit 23 are realized by the processor 20 executing an information processing program stored in the storage unit 30. However, the acquisition unit 21 to the output unit 23 may be composed of a dedicated hardware circuit such as an ASIC.

[0047] The acquisition unit 21 acquires the first structure information D1 and the target element information. The first structure information D1 is information indicating a first structure composed of a plurality of atoms represented by one or more elements. The first structure is, for example, the structure of a unit cell. The first structure will be described later. The target element information is information indicating a target element that is a target for representing the existence probability in the unit cell corresponding to the first structure. Examples of the target element include Li (lithium), F (fluorine), and the like.

[0048] The processing unit 22 uses the first structure information D1 and the target element information to generate first atomic distribution information D10 indicating the distribution of target atoms, which are one or more atoms represented by the target element, in the unit cell corresponding to the first structure.

[0049] The output unit 23 outputs the first atomic distribution information D10. For example, the output unit 23 outputs the first atomic distribution information D10 to the display unit 40 (liquid crystal monitor). However, this is just an example, and the output unit 23 may output the first atomic distribution information D10 to an external device (such as an external server) not shown, which is connected via a communication network such as the Internet and the information processing system 1.

[0050] The output unit 23 may output the first atomic distribution information D10 and / or the second atomic distribution information D20 in a format that can be used by an external program, not limited to display on the display unit 40. For example, the output unit 23 may transmit a list of the existence probabilities of the target atoms associated with identifiers for identifying each atomic site to an external device via an application programming interface. The output unit 23 may output the existence probability as a file associated with the crystal structure data, and the file may be in a format that can be directly input into subsequent analysis processing.

[0051] Subsequently, the flow of the process executed by the information processing system 1 will be described. FIG. 2 is a flowchart showing an example of the process executed by the information processing system 1.

[0052] (Step SP1) In step SP1, the acquisition unit 21 acquires the first structural information D1 and the target element information. Specifically, the acquisition unit 21 displays the information input screen 100 on the display unit 40. The information input screen 100 is a screen for receiving various types of information from the user. In other words, the information input screen 100 functions as a user interface.

[0053] Figure 3 shows an example of the information input screen 100. As shown in Figure 3, the information input screen 100 includes a structure information input field 101, a target element information input field 102, an atom number information input field 103, a temperature information input field 104, and a distribution prediction button 105.

[0054] The information input screen 100 may include a setting item to specify whether or not to calculate the probability distribution of the target atom's existence, or a setting item to specify whether or not to output the probability of existence for each atomic site.

[0055] The structural information input field 101 is used to input first structural information D1, which represents a first structure composed of multiple atoms represented by one or more elements. For example, the crystal structure of a battery material that makes up a battery such as a lithium-ion battery may correspond to the first structure.

[0056] Figure 4 shows an example of first structural information D1. More specifically, Figure 4 is an example of first structural information D1 represented by text written according to the Common Crystallographic Data Format (CIF). The first structural information D1 shown in Figure 4 includes information that identifies the structure of the unit cell, which is the smallest unit of a crystal structure. Specifically, it includes the lengths of the a-axis, b-axis, and c-axis, which are virtual axes for describing the crystal structure; the magnitude of the angle α between the a-axis and b-axis; the magnitude of the angle β between the b-axis and c-axis; the magnitude of the angle γ between the c-axis and a-axis; space group information indicating the space group to which the crystal structure belongs; and so on. Although not shown in the illustration, first structural information D1 also includes information indicating one or more atoms contained in the first structure. For example, information indicating that the first structure contains multiple atoms such as lithium atoms, oxygen atoms, and nitrogen atoms is included in first structural information D1.

[0057] Returning to Figure 3, the target element information is entered in the target element information input field 102. Target element information is information that specifies the target element. In the example shown in Figure 3, Li (lithium) is specified as the target element. Hereafter, one or more atoms represented by the target element will be called target atoms. In the example shown in Figure 3, lithium atoms are the target atoms.

[0058] As described above, the first structure is a structure composed of multiple atoms represented by one or more elements. Here, the target element may or may not be included in the one or more elements mentioned above. As an example, let's assume that the first structure corresponds to a unit cell contained in the positive electrode material of a battery, and the target atom represented by the target element is a lithium atom. In this case, since the lithium atom constitutes the first structure, the target element is included in the one or more elements mentioned above. In this case, the target element is removed from the first structure when the battery is charged. If the first structure is a negative electrode material (for example, graphite made only of element C), and the target atom is a lithium atom, the target atom is placed (inserted) into the first structure when the battery is charged, but since the original first structure does not contain the target atom, it can be said that the target element is not included in the one or more elements.

[0059] In other words, the first structure may or may not contain the target atom. If the first structure contains the target atom, it is composed of multiple atoms represented by two or more elements. If the first structure does not contain the target atom, it is composed of multiple atoms represented by one or more elements.

[0060] When entering target element information in the target element information input field 102 in Figure 3, you may specify an element that is not included in one or more elements constituting the first structure as the target element, or you may specify an element that is included in one or more elements constituting the first structure as the target element.

[0061] In this embodiment, elements representing atoms that move within the battery to realize the battery's charging and discharging process, such as lithium atoms and fluorine atoms, are designated as target elements. In other words, elements representing atoms with relatively high mobility are designated as target elements. However, this is just one example, and various elements can be designated as target elements.

[0062] In the following, atoms that do not correspond to the target atom among the one or more elements constituting the first structure are referred to as non-target atoms. In this embodiment, atoms with relatively low mobility correspond to non-target atoms. Examples of non-target atoms include oxygen atoms and nitrogen atoms.

[0063] Note that there may be multiple types of target atoms. For example, two types of atoms, such as lithium atoms and fluorine atoms, may be designated as target atoms. That is, the user may enter target element information indicating two types of elements, "Li" and "F", in the target element information input field 102.

[0064] The atom number information input field 103 is used to input target atom number information, which specifies the number of target atoms per unit cell. In other words, the atom number information input field 103 is used to input target atom number information, which specifies how many target atoms there are in one unit cell. In the example shown in Figure 3, target atom number information is entered to indicate that there are 8 target atoms per unit cell. Note that inputting target atom number information is not mandatory. That is, the acquisition unit 21 does not necessarily need to acquire target atom number information via the atom number information input field 103. If target atom number information is not entered in the atom number information input field 103, the acquisition unit 21 may acquire a standard value for the number of target atoms from the storage unit 30. The standard value for the number of target atoms is a fixed value that is set in advance by the user or the like.

[0065] The temperature information input field 104 is used to input information indicating the temperature of the space in which the first structure exists. In the example shown in Figure 3, temperature information indicating that the temperature of the space in which the first structure exists is 300 degrees Celsius is entered. Note that inputting temperature information is not mandatory. That is, the acquisition unit 21 does not necessarily need to acquire temperature information via the temperature information input field 104. If temperature information is not entered in the temperature information input field 104, the acquisition unit 21 may acquire a standard temperature value from the storage unit 30. The standard temperature value is a fixed value that has been set in advance by the user or the like.

[0066] The distribution prediction button 105 is a button that commands the processing unit 22 to start the process of generating the first atomic distribution information D10 based on various information entered into the information input screen 100. Specifically, when the acquisition unit 21 detects that the distribution prediction button 105 has been selected by the user, it inputs various information received from the user via the information input screen 100 to the processing unit 22. The processing unit 22 executes the processes from step SP2 onward based on the various information received from the acquisition unit 21.

[0067] (Step SP2) Returning to Figure 2, in step SP2, the processing unit 22 generates second structure information D2. Second structure information D2 is information indicating the second structure, which is the structure obtained by removing the target atom from the first structure. As described above, in this embodiment, an atom with relatively high mobility is designated as the target atom. Therefore, the second structure, which is the structure obtained by removing the target atom, is a structure composed of atoms with relatively low mobility. In other words, the second structure is a structure that corresponds to the framework in the crystal structure. Note that if the target element is not included in the elements that make up the first structure, this process is skipped. If the target element is not included in the elements that make up the first structure, the second structure can be said to be a structure in which the target atom is not placed in the first structure.

[0068] The processing unit 22 generates second structure information D2 based on first structure information D1 and target element information. Figure 5 is a schematic diagram showing the process of generating second structure information D2, which represents the second structure, based on first structure information D1, which represents the first structure, and target element information, which represents the target atom. The process performed in step SP2 will be explained with reference to Figure 5.

[0069] The processing unit 22 first generates a model representing the first structure. Hereinafter, the model representing the first structure will be referred to as the first structural model M1. The first structural model M1 is generated by the processing unit 22 constructing a crystal structure based on the first structural information D1 shown in Figure 4, for example. That is, the first structural model M1 is a model that reproduces the structure of the unit cell shown by the first structural information D1. The first structural model M1 includes target atom icons T and non-target atom icons NT. The target atom icons T correspond to target atoms (such as lithium atoms) included in the first structure. The non-target atom icons NT correspond to non-target atoms (such as oxygen atoms) included in the first structure. The first structural model M1 is an example of the first structural information D1.

[0070] The processing unit 22, which generated the first structural model M1, then removes the target atom icon T from the first structural model M1. As a result, the processing unit 22 generates the second structural model M2. That is, the second structural model M2 is a model that reproduces the structure obtained by removing the target atom from the unit cell structure shown in the first structural information D1. The second structural model M2 is an example of the second structural information D2.

[0071] (Step SP3) Returning to Figure 2, in step SP3, the processing unit 22 generates atomic site information. An atomic site is one or more spaces that are candidates for the position where the target atom is placed. In other words, an atomic site is one or more spaces where the target atom may exist. The processing unit 22 according to this embodiment generates atomic site information based on the second structure information D2. More specifically, atomic site information is generated by predicting the position of the atomic site based on the arrangement of one or more remaining atoms (e.g., oxygen atoms, cobalt atoms, etc.) left in the second structure.

[0072] Figure 6 schematically illustrates the process of generating atomic site information based on the second structural information D2. As shown in Figure 6, the second structural model M2 is composed of multiple polyhedra with non-target atomic icons NT as vertices. For example, the second structural model M2 includes a polyhedron PO formed by the first non-target atomic icon NT1, the second non-target atomic icon NT2, the third non-target atomic icon NT3, the fourth non-target atomic icon NT4, and the fifth non-target atomic icon NT5. In addition, the second structural model M2 is composed of various polyhedra with various non-target atomic icons NT as vertices. The processing unit 22 in this embodiment treats the centroid of each of these multiple polyhedra as an atomic site. The processing unit 22 predicts the position of the atomic sites using existing methods for calculating the centroid of each of the multiple polyhedra, such as centroidal Voronoi tessellation. This predicts the position of all spatial locations in which the target atom may exist.

[0073] The processing unit 22 generates an atomic site model M3 based on the atomic site prediction results. The atomic site model M3 is a model in which atomic site icons SI are placed at the predicted locations (coordinates) where atomic sites exist. The atomic site model M3 is an example of atomic site information.

[0074] (Step SP4) Returning to Figure 2, in step SP4, the processing unit 22 calculates the atomic density distribution of the target atoms using a probability distribution model. For example, the processing unit 22 takes the target element information, the second structure information D2, and the target number of atoms as input and performs a calculation using DFT / 3D-RISM (an example of a probability distribution model).

[0075] Figure 7 schematically illustrates the process of calculating the atomic density distribution based on target element information, second structure information D2, and target atom number information. As an example, we assume that the target atom number information, which indicates that there are 8 target atoms, has been acquired by the acquisition unit 21. When the processing unit 22 performs calculations using DFT / 3D-RISM, it treats the 8 target atoms as a probability distribution and the second structure indicated by the second structure information D2 as a framework. This allows the calculation of density distribution information showing the atomic density distribution of the target atoms. Figure 7 shows a density distribution model M4, which is colored according to the level of atomic density of the target atoms. Density distribution model M4 is an example of density distribution information. In density distribution model M4, the first region AR1, where the atomic density of the target atoms is very high, is colored red; the second region AR2, where the atomic density of the target atoms is high, is colored yellow; the third region AR3, where the atomic density of the target atoms is low, is colored green; and the fourth region AR4, where the atomic density of the target atoms is very low, is colored blue.

[0076] Here, we have described an example using DFT / 3D-RISM as a probability distribution model, but this is just one example. The processing unit 22 does not necessarily have to calculate the atomic density distribution using DFT / 3D-RISM. The processing unit 22 may calculate it using methods based on model equations such as the Bond-Valence Sum method, methods based on geometry such as structural topology, or methods based on computational science such as molecular dynamics calculations.

[0077] The atomic density distribution and first atomic distribution information calculated in this embodiment are calculated with the understanding that they will be used as input data in subsequent material property simulations, structure candidate generation, or conduction path analysis. In other words, this process functions as a pre-processing step in the material analysis process.

[0078] (Step SP5) Returning to Figure 2, in step SP5, the processing unit 22 generates first atomic distribution information D10. The first atomic distribution information D10 is information that shows the distribution of the probability of existence of the target atoms. The processing unit 22 generates the first atomic distribution information D10 based on the atomic site information and the atomic density distribution information.

[0079] Figure 8 is a schematic diagram illustrating the process of generating first atomic distribution information D10 based on atomic site information and atomic density distribution information. The processing unit 22 in this embodiment calculates the probability that a target atom exists at an atomic site by integrating the atomic density distribution of target atoms around that atomic site. Specifically, first the processing unit 22 places a virtual sphere at a predetermined atomic site icon SI in the superimposed model. For example, a virtual sphere with a predetermined radius is placed in the superimposed model with the centroid of the first atomic site icon SI1 shown in atomic site model M3 in Figure 8 as its center. Next, the processing unit 22 superimposes a density distribution model M4 onto the atomic site model M3 shown in Figure 8. Then, the processing unit 22 integrates the atomic density distribution of target atoms inside the sphere. The processing unit 22 then treats the value calculated as a result of the integration as the probability of the existence of a target atom at the atomic site corresponding to the first atomic site icon SI1. The processing unit 22 performs the same process as described above for all atomic sites. In other words, the processing unit 22 places a virtual sphere at each atomic site icon SI and integrates the atomic density distribution of the target atom within each sphere. Based on this, the processing unit 22 calculates the probability of the target atom's presence at all atomic sites. That is, the processing unit 22 calculates the probability of the target atom's presence for each of the one or more atomic sites. However, this is just one example, and the method for calculating the probability of the target atom's presence can be changed as appropriate.

[0080] The processing unit 22, having calculated the probability of the target atom's presence at all atomic sites, generates first atomic distribution information D10, which shows the distribution of the target atom's presence probability. Figure 8 shows a first atomic distribution model M10, which is an example of the first atomic distribution information D10. The first atomic distribution model M10 includes a probability icon IC. The probability icon IC includes a numerical value indicating the probability of the target atom's presence at the atomic site where the probability icon IC is located. For example, the first probability icon IC1 shown in Figure 8 has the numerical value "0.88" superimposed on it. This indicates that there is an 88% probability that the target atom is present at the atomic site where the first probability icon IC1 is located. A pie chart is superimposed on the first probability icon IC1. The pie chart includes a first part R1 and a second part R2. The first part R1 is colored, for example, green, and the second part R2 is colored, for example, white. The proportion occupied by the first part R1 in the pie chart corresponds to the probability (88%) that the target atom is present at the atomic site where the first probability icon IC1 is located. In the pie chart, the proportion occupied by the second part R2 corresponds to the probability (12%) that the target atom is not present at the atomic site where the first probability icon IC1 is located. The first atomic distribution model M10 is an example of the first atomic distribution information D10.

[0081] The integral of the probability of a target atom's presence at all atomic sites will match the number of target atoms specified by the user beforehand. For example, if the user has entered information indicating that there are 8 target atoms, the integral of the probability of a target atom's presence at all atomic sites will be 8.

[0082] (Step SP6) Returning to Figure 2, in step SP6, the processing unit 22 generates second atomic distribution information D20. Second atomic distribution information D20 is information that shows the distribution of the probability of existence of target atoms in the second structure. The processing unit 22 according to this embodiment generates second atomic distribution information D20 based on atomic site information, atomic density distribution information, and second structure information D2. The processing unit 22 generates second atomic distribution information D20 using this first atomic distribution information D10 and second structure information D2.

[0083] Figure 9 schematically shows the process of generating second atomic distribution information D20 based on atomic site information, atomic density distribution information, and second structure information D2. First, the processing unit 22 generates first atomic distribution information D10 based on atomic site information and atomic density distribution information. The method for generating first atomic distribution information D10 has already been explained in step SP5, so a detailed explanation is omitted. Here, it is assumed that the first atomic distribution model M10 shown in Figure 8 has been generated as first atomic distribution information D10. Next, the processing unit 22 generates second atomic distribution information D20 using second structure information D2 and first atomic distribution model M10. For example, the processing unit 22 superimposes the first atomic distribution model M10 onto second structure model M2, which is an example of second structure information D2. That is, it associates the probability of existence of a target atom with each of the one or more atomic sites in the second structure. As a result, the processing unit 22 generates second atomic distribution model M20. Second atomic distribution model M20 is an example of second atomic distribution information D20.

[0084] The second atomic distribution information D20 is output in a format suitable for subsequent automated analysis, as a data structure that explicitly includes the probability of existence for each atomic site. This makes it directly usable for computer processing without the need for human interpretation.

[0085] (Step SP7) Returning to Figure 2, in step SP7, the output unit 23 outputs the first atomic distribution information D10. The output unit 23 according to this embodiment displays an output screen 200 for outputting the first atomic distribution information D10 on the display unit 40. Figure 10 is a diagram showing an example of the output screen 200. The output screen 200 includes an atomic distribution information display field 201, a save button 202, a third structure prediction button 203, a conduction path prediction button 204, and a battery characteristic prediction button 205.

[0086] The first atomic distribution information D10 is displayed in the atomic distribution information display area 201. In the example shown in Figure 10, the second atomic distribution model M20 is displayed in the atomic distribution information display area 201. In step SP7, the output unit 23 may output the second atomic distribution model M20 (an example of the second atomic distribution information D20) as the first atomic distribution information D10 to the atomic distribution information display area 201. However, this is just one example. The output unit 23 may also display the first atomic distribution model M10 shown in Figure 8, etc., in the atomic distribution information display area 201.

[0087] In this embodiment, the atomic distribution information display field 201 outputs an image representing the second atomic distribution model M20. When the output unit 23 outputs an image representing the second atomic distribution model M20, it outputs at least one of the following (a), (b), and (c).

[0088] (a) An image showing a second structure in which one or more atoms (target atoms) represented by the target element are arranged, and the probability value of the existence of the target atom is superimposed on each of the arranged target atoms.

[0089] (b) An image showing a second structure in which target atoms are placed, and a graph indicating the probability of finding each of the placed target atoms is superimposed.

[0090] (c) An image showing a structure in which target atoms are placed in the second structure, each of the placed target atoms is colored, and the probability of the existence of the target atom is indicated by the intensity of the color or transparency of the color applied to the target atom.

[0091] In the example shown in Figure 10, an image satisfying the above conditions (a) and (b) is displayed in the atomic distribution information display area 201. Specifically, the second structural model M2 has probability icons IC corresponding to the target atoms placed on it, and each probability icon IC has a numerical value and a pie chart superimposed on it indicating the probability of the target atom being present at the atomic site where the probability icon IC is located. This image is displayed in the atomic distribution information display area 201.

[0092] The output unit 23 should output an image representing the second atomic distribution model M20 using either a two-dimensional image or a three-dimensional model. A three-dimensional model can also be considered a type of image in a broad sense.

[0093] Note that the second atomic distribution information D20 does not necessarily have to be output as an image. The second atomic distribution information D20 may be output as text written according to the Common Crystallographic Data Format (CIF), or in other data formats. For example, it may be in xyz format or vasp format. Figure 11 shows another example of the second atomic distribution information D20. The second atomic distribution information D20 shown in Figure 11 includes information indicating the label assigned to each atomic site. The second atomic distribution information D20 shown in Figure 11 includes information indicating the probability (occupancy rate) that a target atom is present at each atomic site, and information indicating the coordinates of each atomic site. In addition, the second atomic distribution information D20 shown in Figure 11 includes information indicating the label assigned to the space where non-target atoms are placed, information indicating the coordinates of these spaces, information indicating the probability (occupancy rate) that non-target atoms are placed in these spaces, etc.

[0094] The second atomic distribution information D20 is data that includes the numerical probability of the existence of a target atom associated with each atomic site, and this numerical value is output in file format along with the crystal structure data. For example, the existence probability may be output as the occupancy rate of each atomic site, added to the crystal structure data.

[0095] Returning to Figure 10, the save button 202 is a button for saving the second atomic distribution information D20 displayed in the atomic distribution information display field 201 to the storage unit 30. When the output unit 23 detects that the save button 202 has been selected by the user, it stores the second atomic distribution information D20 displayed in the atomic distribution information display field 201 in a predetermined storage area of ​​the storage unit 30.

[0096] The third structure prediction button 203 is a button used to instruct the information processing system 1 to start the prediction process for the third structure. The third structure is composed of multiple atoms represented by one or more elements, and the target atoms are arranged with a probability corresponding to the probability of existence of the target atom indicated by the second atom distribution information D20.

[0097] In this embodiment, after the first atomic distribution information D10 or the second atomic distribution information D20 is output, a physical property simulation, structure candidate generation, or material evaluation process is executed using the atomic distribution information as input. The third structure prediction process is a step in generating a structural model that will actually be the subject of physical property calculations in material design or material evaluation. A crystal structure in which target atoms are arranged at each atomic site, reflecting the probability of existence of the target atom at each atomic site, is considered to be a realistic crystal structure. As one method for generating this crystal structure, it is conceivable to perform a simulation in which target atoms are arranged at each atomic site according to the probability of existence of the target atom at each atomic site. Specifically, it is conceivable to perform a simulation in which, based on the second structure information D2 and the second atomic distribution information D20, target atoms are arranged at each of the one or more atomic sites formed in the second structure, taking into account the probability of existence of the target atom at each atomic site. For example, suppose there are 10 atomic sites in the second structure, and the probability of a target atom being present at each of these atomic sites is 80%. Here, it is expected that a realistic crystal structure will be generated if a simulation using the above method is performed. Specifically, it is expected that a structure will be generated in which the target atom is located in 8 out of 10 atomic sites. However, due to probability issues, a realistic crystal structure is not always guaranteed to be generated. For example, a crystal structure may be generated in which the target atom is not located in any of the 10 atomic sites. Therefore, the processing unit 22 in this embodiment performs a third structure prediction process in order to generate a realistic crystal structure, that is, a third structure composed of multiple atoms represented by one or more elements, in which the target atom is located with a probability corresponding to the probability of existence of the target atom indicated by the second atomic distribution information D20.

[0098] Figure 12 is a flowchart showing the process for predicting the third structure. When the processing unit 22 detects that the third structure prediction button 203 has been selected, it starts the process shown in Figure 12.

[0099] (Step SP21) In step SP21, the processing unit 22 generates third structural information D30 indicating the third structure. The processing unit 22 according to this embodiment generates third structural information D30 based on target element information, second structural information D2, and first atomic distribution information D10.

[0100] For example, the processing unit 22 generates third structural information D30 using the cluster expansion method. Specifically, the processing unit 22 randomly generates crystal structures using target element information, second structural information D2, and first atomic distribution information D10 as input. By repeating this process, the processing unit 22 generates multiple crystal structures. For example, the processing unit 22 generates 100 crystal structures. Subsequently, the processing unit 22 performs first-principles calculations on each of the randomly generated crystal structures to predict the energy of each crystal structure. This generates pairs of data representing the crystal structure and data representing the energy in that crystal structure. In this example, 100 pairs are generated. Then, the processing unit 22 constructs a machine learning model that predicts energy based on crystal structures by performing machine learning using the above data pairs as training data.

[0101] The processing unit 22, which has constructed a machine learning model, takes the target element information, the second structure information D2, and the first atomic distribution information D10 as input and randomly generates multiple crystal structures again. Then, the processing unit 22 inputs the randomly generated multiple crystal structures into the machine learning model. As a result, the machine learning model outputs the energy of each of the multiple crystal structures. Based on the energy of each crystal structure output by the machine learning model, the processing unit 22 extracts crystal structures with low energy. For example, the processing unit 22 extracts crystal structures whose energy is below a predetermined threshold. The processing unit 22 randomly generates multiple crystal structures, inputs these crystal structures into the machine learning model, has the machine learning model output the energy of each crystal structure, and repeats the process of extracting crystal structures whose energy is below a threshold a predetermined number of times. As a result, a group of crystal structures with low energy is obtained. Then, the processing unit 22 treats the crystal structure with the lowest energy among the group of crystal structures as the third structure. However, this is just an example, and the method for predicting the third structure can be changed as appropriate.

[0102] In this way, by selecting structural candidates using first-atom distribution information, the number of structures subject to physical property calculations can be significantly reduced, improving the efficiency of computational resource usage.

[0103] Figure 13 schematically illustrates the process of generating third structure information D30 based on target element information, second structure information D2, and first atomic distribution information D10. Figure 13 shows a third structure model M30 that visualizes the crystal structure of the third structure. The third structure model M30 is an example of third structure information D30. The third structure model M30 includes target atom icons T corresponding to target atoms and non-target atom icons NT corresponding to non-target atoms. In the third structure model M30, the target atom icons T are placed at each of the one or more atomic sites reproduced on the third structure model M30 with a probability corresponding to the probability of existence of the target atom. If the probability of existence of the target atom at each atomic site in the third structure model M30 is 80%, then on average, the target atom icons T are placed at each atomic site with a probability of 80% for the entire third structure model M30. In other words, the probability of existence of the target atom at each atomic site is reflected in the third structure model M30. Therefore, the third structural model M30 can be said to be a model that reproduces a realistic crystal structure.

[0104] However, in the third structural model M30, the target atom icon T does not necessarily have to be positioned with a probability of perfectly matching 80%, but may be positioned with a probability within the tolerance range.

[0105] Furthermore, the third structural model M30 shown in Figure 13 has a structure in which the volume of the unit cell, which is virtually reproduced by placing target atoms at each atomic site present in the second structural model M2 according to the probability of existence of the target atoms indicated by the first atomic distribution information D10, is multiplied by eight (twice in the a-axis, b-axis, and c-axis directions, respectively).

[0106] (Step SP22) Returning to Figure 12, in step SP22, the output unit 23 outputs the third structure information D30. The output unit 23 displays, for example, the third structure display screen 300 shown in Figure 10 on the display unit 40. The third structure display screen 300 includes a third structure display field 301. In the example shown in Figure 10, the third structure display field 301 displays a third structure model M30, which is an example of the third structure information D30. When the processing unit 22 detects that the save button included in the third structure display screen 300 has been selected by the user, it stores the third structure model M30 displayed in the third structure display field 301 in a predetermined storage area of ​​the storage unit 30.

[0107] Next, the conduction path prediction process will be described. When the processing unit 22 detects that the conduction path prediction button 204 shown in Figure 10 has been selected, it starts the conduction path prediction process to predict the ion conduction path. The processing unit 22 in this embodiment predicts the ion conduction path by predicting the conduction path of the target atom.

[0108] In this embodiment, atomic distribution information is used as a search condition to eliminate unrealistic conduction paths that may arise if the probability of atomic existence is not considered.

[0109] Figure 14 is a diagram illustrating an example of conduction path prediction processing. Figure 14 shows a first plan view 600 that represents the second structural model M2 in a planar manner. The grid-like frame FL included in the first plan view 600 corresponds to a plurality of polyhedra that constitute the second structural model M2, and each of the plurality of spaces enclosed by the grid-like frame FL corresponds to an atomic site. In the example shown in Figure 14, the first plan view 600 includes the first atomic site S1 to the ninth atomic site S9.

[0110] In order to predict the ion conduction path, the processing unit 22 first sets the probability of the target atom's presence at each atomic site based on the first atomic distribution information D10 or the second atomic distribution information D20. In the example shown in Figure 14, the probability of the target atom's presence at the first atomic site S1 to the third atomic site S3 is set to 80%, the probability of the target atom's presence at the fourth atomic site S4 to the sixth atomic site S6 is set to 20%, and the probability of the target atom's presence at the seventh atomic site S7 to the ninth atomic site S9 is set to 80%.

[0111] Next, the processing unit 22 determines the starting point for the target atom's movement. For example, the processing unit 22 may determine the atomic site with the highest probability of the target atom's existence as the starting point. However, this is just one example. The processing unit 22 may also determine the starting point randomly. Alternatively, the processing unit 22 may extract atomic sites where the probability of the target atom's existence is above a threshold (e.g., 80%), randomly select one atomic site from the extracted sites, and determine that selected atomic site as the starting point. For example, the processing unit 22 may determine the first atomic site S1 as the starting point.

[0112] Next, the processing unit 22 checks the probability of the target atom being present at an atomic site adjacent to the starting point of movement (hereinafter referred to as an adjacent site). In this example, the second atomic site S2 and the fourth atomic site S4 are adjacent sites. Therefore, the processing unit 22 checks the probability of the target atom being present at the second atomic site S2 and the fourth atomic site S4, respectively.

[0113] Next, the processing unit 22 determines whether there are any adjacent sites where the probability of the target atom being present exceeds a standard value (for example, 80%). If there are adjacent sites that satisfy the above conditions, the processing unit 22 treats those adjacent sites as the atomic sites to which the target atom will move (hereinafter referred to as the destination sites). As described above, the probability of the target atom being present at the second atomic site S2 is 80%, and the probability of the target atom being present at the fourth atomic site S4 is 20%. Therefore, in the example shown in Figure 14, the second atomic site S2 is the destination site. If there are no destination sites, the processing unit 22 determines the first atomic site S1 as the end point of the movement.

[0114] Next, the processing unit 22 determines whether an atomic site where the probability of the target atom's existence exceeds a reference value (e.g., 80%) is adjacent to the destination site. In other words, the processing unit 22 searches for the next destination site. In the example shown in Figure 14, the third atomic site S3 corresponds to the next destination site. The processing unit 22 repeats the process of searching for the next destination site until there are no more suitable atomic sites available.

[0115] As shown in Figure 14, the third atomic site S3 is adjacent to the sixth atomic site S6, but the probability of the target atom being present at the sixth atomic site S6 is 20%. In this case, the processing unit 22 determines that there are no suitable atomic sites as the next destination site for the third destination site and terminates the process of searching for the next destination site. At this time, the processing unit 22 determines the third atomic site S3 as the destination site. The destination site is set such that the x, y, and z components of the distance between the start and end points are greater than or equal to the x, y, and z components of the lattice constants of the unit cell.

[0116] Next, the processing unit 22 identifies an ion conduction path based on the atomic site determined as the starting point of movement, the atomic site determined as the ending point of movement, and the atomic sites that were traversed from the starting point to the ending point of movement. In the example shown in Figure 14, the processing unit 22 identifies a path 551 from the first atomic site S1 to the third atomic site S3 via the second atomic site S2 as the ion conduction path.

[0117] Figure 15 is a diagram illustrating another example of conduction path prediction processing. Figure 15 shows a second plan view 700 schematically representing the second structural model M2. The grid-like frame FL included in the second plan view 700 corresponds to a plurality of polyhedra that constitute the second structural model M2, and each of the plurality of spaces formed by the grid-like frame FL corresponds to an atomic site. In the example shown in Figure 15, the second plan view 700 includes the 11th atomic site S11 to the 19th atomic site S19. In the example shown in Figure 15, the probability of finding the target atom at the 11th atomic site S11 is set to 80%, the probability of finding the target atom at the 12th atomic site S12 and the 13th atomic site S13 is set to 20%, the probability of finding the target atom at the 14th atomic site S14 to the 16th atomic site S16 is set to 80%, the probability of finding the target atom at the 17th atomic site S17 and the 18th atomic site S18 is set to 20%, and the probability of finding the target atom at the 19th atomic site S19 is set to 80%.

[0118] The processing unit 22 identifies the ion conduction path by determining the starting point of movement, searching for a suitable atomic site as the destination site from among the atomic sites adjacent to the starting point of movement, and searching for a suitable atomic site as the next destination site after the destination site, similar to the example described in Figure 14. In the example shown in Figure 15, the path 552 from the 11th atomic site S11 to the 19th atomic site S19, via the 14th atomic site S14, the 15th atomic site S15, and the 16th atomic site S16, is identified as the ion conduction path.

[0119] The processing unit 22, which has identified the ion conduction path, inputs ion conduction path information indicating the ion conduction path to the output unit 23. The output unit 23 then outputs the ion conduction path information to the display unit 40. For example, the output unit 23 outputs the conduction path display screen 400 shown in Figure 10 to the display unit 40. As shown in Figure 10, the conduction path display screen 400 includes a conduction path display field 401. The output unit 23 displays the ion conduction path information in the conduction path display field 401. In the example shown in Figure 10, a conduction path model M40 (an example of ion conduction path information) in which the ion conduction path is visualized is displayed in the conduction path display field 401.

[0120] Furthermore, if there are multiple suitable atomic sites as destination sites, the processing unit 22 identifies multiple ion conduction paths. For example, if the probability of the target atom being present at the 18th atomic site S18 shown in Figure 15 is set to 80%, then the 16th atomic site S16 and the 18th atomic site S18 are candidates for the next destination site after the 15th atomic site S15. In this case, the processing unit 22 identifies not only the path 552 from the 15th atomic site S15 to the 19th atomic site S19 via the 16th atomic site S16, but also another path from the 15th atomic site S15 to the 19th atomic site S19 via the 18th atomic site S18 as an ion conduction path. In this case, the conduction path display area 401 shown in Figure 10 should display a conduction path model M40 in which each of the multiple ion conduction paths is visualized.

[0121] The property prediction process according to this embodiment is a process for comparing and evaluating multiple material candidates in material design. Next, the property prediction process for predicting the properties of the first structure will be described.

[0122] The following describes an example of characteristic prediction processing, specifically the prediction of the characteristics of an object (for example, a lithium-ion battery) that includes a first structure. When the processing unit 22 detects that the battery characteristic prediction button 205 shown in Figure 10 has been selected by the user, it starts the characteristic prediction processing. The processing unit 22 in this embodiment predicts the charge-discharge characteristics, which are an example of characteristics.

[0123] Figure 16 is a diagram illustrating the characteristic prediction process. Figure 16 shows a third plan view 800 that represents the second structural model M2 in a planar manner. The grid-like frame FL included in the third plan view 800 corresponds to a plurality of polyhedra that constitute the second structural model M2, and each of the plurality of spaces formed by the grid-like frame FL corresponds to an atomic site. In the example shown in Figure 16, the third plan view 800 includes the 21st atomic site S21 to the 29th atomic site S29.

[0124] First, the processing unit 22 sets the probability of existence of the target atom for each of the 21st atomic site S21 to the 29th atomic site S29 based on the first atomic distribution information D10 or the second atomic distribution information D20. In the example shown in Figure 16, the probability of existence of the target atom at the 21st atomic site S21 (hereinafter simply referred to as the probability of existence) is set to 90%, at the 22nd atomic site S22 to 54%, at the 23rd atomic site S23 to 88%, at the 24th atomic site S24 to 60%, at the 25th atomic site S25 to 80%, at the 26th atomic site S26 to 50%, at the 27th atomic site S27 to 75%, at the 28th atomic site S28 to 30%, and at the 29th atomic site S29 to 70%.

[0125] Next, the processing unit 22 virtually generates a structure in which no target atoms are placed at any of the atomic sites. For example, the processing unit 22 generates a structure in which no target atom icon T is placed at each of the 21st atomic site S21 to the 29th atomic site S29 in the third plan view 800. In other words, the processing unit 22 reproduces the fully discharged state of the battery.

[0126] Next, the processing unit 22 generates a structural model by placing (inserting) the target atom icon T at the atomic site with the highest probability of the target atom's existence among all atomic sites. For example, the processing unit 22 generates a structural model by placing the target atom icon T at the 21st atomic site S21. Hereinafter, the structural model in which the target atom icon T is placed at the atomic site with the highest probability of the target atom's existence will be referred to as the first model.

[0127] Next, the processing unit 22 predicts the energy of the first model by performing a physical property simulation using the first model as input. The processing unit 22 stores the predicted energy of the first model in the storage unit 30.

[0128] Next, the processing unit 22 places the target atom icon T at the atomic site with the highest probability of being present among the remaining atomic sites (the 22nd atomic site S22 to the 29th atomic site S29). As a result, the processing unit 22 generates a structural model with a different number of target atoms than the first model. For example, the processing unit 22 generates a second model by placing (inserting) the target atom icon T at the 23rd atomic site S23.

[0129] Next, the processing unit 22 predicts the energy of the second model by performing a physical property simulation using the second model as input. The processing unit 22 stores the predicted energy of the second model in the storage unit 30.

[0130] The processing unit 22 repeatedly generates a structural model by placing the target atom icon T at the atomic site with the highest probability of the target atom's existence among the remaining atomic sites. This process is repeated until the target atom icon T is placed at all atomic sites. As a result, multiple structural models (for example, the third to ninth models) are generated. Figure 16 schematically shows the process of generating the sixth model. In other words, Figure 16 schematically shows the process of generating the sixth model by placing the target atom icon T at the 24th atomic site S24, which is the atomic site with the sixth highest probability of the target atom's existence. Each time a new structural model is generated, the processing unit 22 performs a physical property simulation using the newly generated structural model as input and predicts the energy of the newly generated structural model. The processing unit 22 stores the predicted energy of the newly generated structural model in the storage unit 30.

[0131] As described above, the processing unit 22 according to this embodiment generates multiple structural models by gradually increasing the number of target atom icons T placed at atomic sites. This reproduces the process of a battery being charged. Moreover, since the processing unit 22 according to this embodiment generates each structural model by arranging (inserting) the target atom icons T in order of decreasing probability of the target atom's presence, a realistic charging process is accurately reproduced. In other words, the process in which target atoms move to the negative or positive electrode side of the battery and energy is stored in the positive or negative electrode is accurately reproduced.

[0132] In this way, by generating a structural model based on atomic distribution information and performing material property simulations, it is possible to predict practical material properties with high accuracy while reducing trial-and-error calculations.

[0133] Next, the processing unit 22 calculates the energy difference between the structural models. For example, the processing unit 22 calculates the energy difference between the first model and the second model based on the energy of the first model and the energy of the second model. Similarly, the processing unit 22 calculates the energy difference between the energy of the second model and the energy of the third model. The processing unit 22 repeats the above process until all energy differences have been calculated.

[0134] Next, the processing unit 22 calculates the voltage of each structural model based on all the energy differences. That is, it calculates the voltages of the first to ninth models.

[0135] Next, the processing unit 22 generates a charging curve by plotting the voltages of the first to ninth models in coordinate space. That is, it generates a charging curve by sequentially plotting the voltages of the first model, the second model, the third model, etc., in coordinate space. This charging curve shows a charging characteristic, which is an example of the characteristics. In this embodiment, since the charging curve is generated after accurately reproducing the actual charging process, a highly accurate charging curve is generated.

[0136] Next, the processing unit 22 virtually generates a structure in which target atoms are placed at all atomic sites. For example, it generates a structure in which target atom icons T are placed at each of the 21st atomic site S21 to the 29th atomic site S29 shown in Figure 16. In other words, the processing unit 22 reproduces the fully charged state of the battery.

[0137] Next, the processing unit 22 generates a structural model by removing the target atom icon T from the atomic site with the lowest probability of the target atom's existence. For example, the processing unit 22 generates a structural model by removing the target atom from the 28th atomic site S28. Hereinafter, the structural model from which the target atom icon T has been removed from the atomic site with the lowest probability of the target atom's existence will be referred to as the 11th model.

[0138] Next, the processing unit 22 predicts the energy of the 11th model by performing a physical property simulation using the 11th model as input. The processing unit 22 stores the predicted energy of the 11th model in the storage unit 30.

[0139] Next, the processing unit 22 removes the target atom from the atomic site with the lowest probability of being present among the remaining atomic sites (in this case, the 21st atomic site S21 to the 27th atomic site S27 and the 29th atomic site S29). For example, the processing unit 22 removes the target atom from the 26th atomic site S26. As a result, the processing unit 22 generates a 12th model in which the number of target atom icons T differs from that of the 11th model.

[0140] Next, the processing unit 22 predicts the energy of the 12th model by performing a physical property simulation using the 12th model as input. The processing unit 22 stores the predicted energy of the 12th model in the storage unit 30.

[0141] The processing unit 22 repeatedly generates a structural model by removing the target atom icon T from the atomic site with the lowest probability of being present among the remaining atomic sites. This process may be repeated until the target atom icon T is removed from all atomic sites, or until the number of atoms falls below a certain value. This generates multiple structural models (for example, the 13th to the 19th model). Each time a new structural model is generated, the processing unit 22 performs a physical property simulation using the structural model as input and predicts the energy of the structural model. The processing unit 22 stores the prediction results in the storage unit 30.

[0142] As described above, the processing unit 22 in this embodiment generates multiple structural models by gradually decreasing the number of target atom icons T placed at atomic sites. This reproduces the process of battery discharge. Moreover, since the processing unit 22 in this embodiment generates each structural model by removing target atom icons T in order of decreasing probability of the presence of the target atom, a realistic discharge process is accurately reproduced. In other words, the process in which target atoms move to the outside from the negative or positive electrode side of the battery and energy is released is accurately reproduced.

[0143] Next, the processing unit 22 calculates the energy difference between the structural models and calculates the voltage of each structural model based on the energy difference.

[0144] Next, the processing unit 22 generates a discharge curve by plotting the voltages of the 11th to 19th models in coordinate space. That is, the discharge curve is generated by sequentially plotting the voltages of the 11th model, the 12th model, the 13th model, etc., in coordinate space. This discharge curve shows the discharge characteristics, which are an example of the characteristics. In this embodiment, since the discharge curve is generated after accurately reproducing the actual discharge process, a highly accurate discharge curve is generated.

[0145] Next, the processing unit 22 inputs the above-mentioned charging curve and discharge curve (an example of characteristic information) to the output unit 23.

[0146] Next, the output unit 23 outputs characteristic information. For example, the output unit 23 outputs the characteristic information display screen 500 shown in Figure 10 to the display unit 40. The characteristic information display screen 500 includes a characteristic information display area 501. In the example shown in Figure 10, the characteristic information display area 501 displays a characteristic graph 510 that reflects the above-mentioned charging curve and discharge curve. The horizontal axis of the characteristic graph 510 represents capacity, and the vertical axis represents voltage.

[0147] According to the information processing system 1 described above, a first atomic distribution information D10 is output, which shows the distribution of the probability of existence of the target atom. This first atomic distribution information D10 can be used when predicting the properties of the first structure. For example, suppose the target atom is a lithium atom and the first structure is a battery material for a lithium-ion battery. In this case, it becomes possible to predict the properties of the battery material after considering the distribution of the probability of existence of lithium atoms in the battery material.

[0148] According to the information processing system 1 of this embodiment, second atom distribution information D20, which shows the distribution of the probability of existence of the target atom in the second structure, is output. This second atom distribution information D20 can be used, for example, when evaluating the properties of the first structure, similar to the first atom distribution information D10 described above.

[0149] According to the information processing system 1 of this embodiment, the positions of one or more atomic sites that are candidates for the position where the target atom will be placed are predicted based on the arrangement of the remaining atoms left in the second structure. Therefore, the positions of one or more atomic sites can be predicted more accurately compared to the case in which the positions of one or more atomic sites are predicted without using the arrangement of the remaining atoms.

[0150] According to the information processing system 1 of this embodiment, it is possible to generate second atomic distribution information D20 in which the probability of the presence of a target atom is superimposed on each of the one or more atomic sites in the second structure. For example, it is possible to generate a second atomic distribution model M20 as shown in Figure 10, etc. In other words, it is possible to generate second atomic distribution information D20 that clarifies which atomic sites have what probability of containing the target atom.

[0151] According to the information processing system 1 of this embodiment, the probability of a target atom being present at each of one or more atomic sites is calculated using a probability distribution model. This makes it possible to calculate the probability of a target atom being present at each atomic site more accurately.

[0152] According to the information processing system 1 of this embodiment, at least one of the following images representing the second atomic distribution information D20 is output: (a) an image in which the probability of existence of target atoms for each atomic site is visualized numerically, (b) an image in which the probability of existence of target atoms for each atomic site is visualized graphically, or (c) an image in which the probability of existence of target atoms for each atomic site is visualized by the intensity of color or transparency. This makes it possible for the user to intuitively understand the probability of existence of target atoms for each atomic site.

[0153] According to the information processing system 1 of this embodiment, it is possible to generate third structure information D30 that shows a third structure in which target atoms are arranged with a probability corresponding to the probability of existence of the target atoms shown in the second atomic distribution information D20. In other words, according to the information processing system 1, it is possible to generate a realistic crystal structure in which target atoms are arranged at each atomic site in a manner that reflects the probability of existence of the target atoms at each atomic site.

[0154] According to the information processing system 1 of this embodiment, the ion conduction path is predicted based on the second atomic distribution information D20. That is, the ion conduction path is predicted after considering the distribution of the probability of existence of the target atoms. Therefore, the ion conduction path can be predicted more accurately compared to the case where the ion conduction path is predicted without using the second atomic distribution information D20.

[0155] According to the information processing system 1 of this embodiment, the characteristics of the first structure are predicted based on the second atomic distribution information D20. That is, the characteristics of the first structure are predicted after considering the distribution of the probability of existence of the target atoms. Therefore, compared to the case where the characteristics of the first structure are predicted without using the second atomic distribution information D20, the characteristics of the first structure can be predicted more accurately.

[0156] This disclosure may be modified as follows:

[0157] (1) In Embodiment 1, an example was described in which the information processing system 1 is used to predict the properties of a battery material, but this is just one example. The information processing system 1 can be used to predict the properties of various objects other than battery materials. For example, it can be used to predict the distribution of the probability of existence of target atoms contained in an object other than a battery material, and then predict the properties of the object other than a battery material based on this distribution of probability.

[0158] (Other) (1) In this disclosure, “at least one selected from the group consisting of A1, A2, ..., and An” may be interpreted as “A1, A2, ..., An, or any combination of A1, A2, ..., An,” where n is an integer greater than or equal to 2.

[0159] For example, at least one of A1, A2, and A3 may be interpreted as "A1", "A2", "A3", "A1 and A2", "A1 and A3", "A2 and A3", or "A1, A2, and A3".

[0160] This disclosure is useful for predicting the probability distribution of the presence of an atom.

[0161] 1: Information processing system 10: Input unit 20: Processor 21: Acquisition unit 22: Processing unit 23: Output unit 30: Storage unit 40: Display unit 100: Information input screen 200: Output screen 300: Third structure display screen 301: Third structure display field 400: Conduction path display screen 401: Conduction path display field 500: Characteristic information display screen 501: Characteristic information display field 510: Characteristic graph (example of characteristic information) D1: First structure information D10: First atomic distribution information D2: Second structure information D20: Second atomic distribution information D30: Third structure information M1: First structure model (example of first structure information) M10: First atomic distribution model (example of first atomic distribution information) M2 : Second structural model (an example of second structural information) M20: Second atomic distribution model (an example of second atomic distribution information) M3: Atomic site model (an example of atomic site information) M30: Third structural model (an example of third structural information) M4: Density distribution model M40: Conduction path model (an example of ion conduction path information) NT: Non-target atom icon SI: Atomic site icon T: Target atom icon

Claims

1. An information processing method performed by a computer, comprising: acquiring first structure information indicating a first structure composed of multiple atoms, and target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure; generating first atom distribution information indicating the distribution of the probability of existence of one or more target atoms represented by the target element in a unit cell corresponding to the first structure, using the first structure information and the target element information; and outputting the first atom distribution information.

2. The information processing method according to claim 1, wherein the first structure is composed of a plurality of atoms represented by two or more elements, the target element is one of the two or more elements, and further comprises: generating second structure information showing a second structure obtained by removing the target atom from the first structure; generating second atom distribution information showing the distribution of the probability of existence of the target atom in the second structure using the second structure information and the target element information; and outputting the second atom distribution information.

3. The information processing method according to claim 1, comprising: generating second structure information showing a second structure in which the target atoms are not arranged in the first structure; generating second atom distribution information showing the distribution of the probability of the target atoms being present in the second structure using the second structure information and the target element information; and outputting the second atom distribution information.

4. The information processing method according to claim 2, further comprising generating atomic site information by predicting the positions of one or more atomic sites that are candidates for the positions in which the target atom is located, based on the arrangement of the remaining atoms left in the second structure.

5. The information processing method according to claim 4, wherein generating the second atomic distribution information includes associating the probability of existence of the target atom with each of the one or more atomic sites in the second structure, based on the second structure information, the target element information, and the atomic site information.

6. The information processing method according to claim 4 or 5, wherein generating the second atomic distribution information includes calculating the probability of existence of the target atom for each of the one or more atomic sites using a probability distribution model.

7. Outputting the second atomic distribution information includes outputting an image representing the second atomic distribution information, wherein the second atomic distribution information represented by the image includes at least one of the following: (a) information showing a structure in which the target atoms are arranged in the second structure, and a value indicating the probability of existence of each of the arranged target atoms is superimposed on each of the arranged target atoms; (b) information showing a structure in which the target atoms are arranged in the second structure, and a graph indicating the probability of existence of each of the arranged target atoms is superimposed on each of the arranged target atoms; (c) information showing a structure in which the target atoms are arranged in the second structure, and each of the target atoms is colored, and the probability of existence of each of the target atoms is represented by the intensity of the color or the transparency of the color.

8. The information processing method according to claim 2, further comprising: generating third structure information, which is a third structure composed of a plurality of atoms represented by one or more elements, and in which the target atoms are arranged in a probability corresponding to the probability of existence of the target atoms indicated by the second atom distribution information, using the target element information, the second structure information, and the first atom distribution information; and outputting the third structure information.

9. The information processing method according to claim 2, further comprising: generating ion conduction path information indicating the ion conduction path of the target atom by performing conduction path prediction processing based on the second atomic distribution information; and outputting the ion conduction path information.

10. The information processing method according to claim 2, further comprising: generating a plurality of structural models in which the number of target atoms differs from each other based on the second atomic distribution information; generating characteristic information indicating the properties of the first structure by performing a physical property simulation using the plurality of structural models as input; and outputting the characteristic information.

11. An information processing system including a processor, wherein the processor performs the following actions: acquires first structure information indicating a first structure composed of a plurality of atoms, and target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure; generates first atom distribution information indicating the distribution of the probability of existence of one or more target atoms represented by the target element in a unit cell corresponding to the first structure, using the first structure information and the target element information; and outputs the first atom distribution information.

12. A program that causes a computer to perform the following actions: acquire first structure information indicating a first structure composed of multiple atoms, and target element information indicating a target element representing the probability of existence in a unit cell corresponding to the first structure; generate first atom distribution information indicating the distribution of the probability of existence of one or more target atoms represented by the target element in a unit cell corresponding to the first structure, using the first structure information and the target element information; and output the first atom distribution information.

Images