Information Processing Method, Information Processing Apparatus, and Information Processing Program

Abstract

To provide an information processing method capable of easily estimating a situation in a prescribed space.SOLUTION: An information processing method relating to a flow regulation body in a predetermined space comprises: a storage step for storing a learning model learning the relationship between information representing the state of the predetermined space, information relating to the flow regulation body included in each small space in a case where a plurality of small spaces included in the predetermined space are defined, and information relating to the flow in that case; a reception step for receiving input of information representing the state of the predetermined space and one-dimensional information indicating information relating to the flow regulation body included in each small space in the case where a plurality of small spaces included in the predetermined space are defined; an estimation step for inputting the one-dimensional information received in the reception step into the learning model and estimating the state of the predetermined space indicated by the one-dimensional information and the flow in the predetermined space in a case where the flow regulation body is arranged; and an output step for outputting information relating to the flow in the predetermined space based on the estimation in the estimation step.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and an information processing program.

Background Art

[0002] Conventionally, in order to obtain a desired result, various simulations are performed and then an actual device is created.

[0003] For example, when creating a silicon rod as described in Patent Document 1, it is necessary to perform simulations on the arrangement of the rods and the arrangement of the gas nozzles for inserting gas into the bell jar. This is because if the appropriate arrangement is not made, the desired result cannot be obtained.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the method of performing a plurality of simulations as described above and setting the conditions that give the most desirable result as the conditions of the actual device, there are problems such as the time required for a plurality of simulations and the possibility that conditions that could give better results than the plurality of simulations may be left out.

[0006] Therefore, the present invention has been made in view of the above problems, and an object thereof is to provide an information processing apparatus that can obtain conditions that are closer to the optimal solution in a shorter time than performing a plurality of simulations.

Means for Solving the Problems

[0007] In order to solve the above problems, an information processing method according to an aspect of the present invention is an information processing method by an information processing apparatus that inputs and outputs information regarding a flow defining body that defines a flow within a predetermined space for a predetermined space, including: a storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow defining body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; a reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow defining body included in each small space when a plurality of small spaces included in the predetermined space are defined; an estimation step of inputting the one-dimensional information received by the reception step into the learning model and estimating the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow defining body indicated by the one-dimensional information are set; and an output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step.

[0008] In the above information processing method, the predetermined space may be the space inside the container of a deposition reactor that deposits a compound based on a raw material on the surface of a core material, the container has ejection holes for ejecting the raw material, the flow defining body may be the ejection holes for ejecting the raw material, and the reception step may receive the number of ejection holes included in a plurality of small spaces.

[0009] In the above information processing method, the information regarding the flow within the predetermined space may be information regarding the heat flux generated around the core material.

[0010] In the above information processing method, the information regarding the flow within the predetermined space may be information regarding the thickness of the compound deposited on the core material.

[0011] In the above information processing method, the predetermined space is the space inside the container of the deposition reactor that deposits a compound based on a raw material on the surface of the core material. The container is provided with an ejection hole for ejecting the raw material and a discharge hole for discharging the gas in the predetermined space. The flow restricting body is the discharge hole, and the reception step may be to receive the number of discharge holes included in a plurality of small spaces.

[0012] In the above information processing method, the predetermined space is the space inside the container filled with powder particles. By supplying gas into the container, the powder particles will flow. The flow restricting body is an obstruction plate provided at a position that obstructs the flow in a predetermined direction within the predetermined space. The air bubbles contained within the predetermined space will split upon contact with the obstruction plate and their diameter will become smaller. The information regarding the flow within the predetermined space may be information regarding at least either the diameter or the number of the air bubbles contained within a measurement region within the predetermined space.

[0013] In the above information processing method, the predetermined space is the space inside the container filled with powder particles, and it has a measurement point for measuring the temperature inside that space. A temperature control means is arranged within the powder particles, and by supplying gas into the container, the powder particles will flow. The flow restricting body is the temperature control means for temperature adjustment, and the information regarding the flow within the predetermined space may be the temperature of the measurement point within the predetermined space.

[0014] In the above information processing method, the predetermined space has a measurement region for measuring at least either the temperature or the concentration of a substance within that space, and it is a step room where gas flows. The flow restricting body is a vent or an opening provided in the predetermined space, and the information regarding the flow within the predetermined space may be information regarding at least either the temperature or the concentration of the substance within the measurement region within the predetermined space.

[0015] In order to solve the above problems, an information processing apparatus according to an aspect of the present invention is an information processing apparatus that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space. The information processing apparatus includes a storage unit that stores a learning model that has learned the relationship between information representing the state of a predetermined space, information regarding the flow regulation bodies included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; a reception unit that receives an input of one-dimensional information indicating information representing the state of a predetermined space and information regarding the flow regulation bodies included in each small space when a plurality of small spaces included in the predetermined space are defined; an estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulation bodies indicated by the one-dimensional information are set; and an output unit that outputs information regarding the flow within the predetermined space based on the estimation by the estimation unit.

[0016] In order to solve the above problems, a communication situation information output program according to an aspect of the present invention causes a computer that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space to realize a reception function that receives an input of one-dimensional information indicating information representing the state of a predetermined space and information regarding the flow regulation bodies included in each small space when a plurality of small spaces included in the predetermined space are defined, an estimation function that inputs the one-dimensional information received by the reception function into a learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulation bodies indicated by the one-dimensional information are set, and an output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function, the computer being capable of accessing a storage unit that stores a learning model that has learned the relationship between information representing the state of a predetermined space, information regarding the flow regulation bodies included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case.

Advantages of the Invention

[0017] An information processing apparatus according to one aspect of the present invention uses a learning model that has learned the relationship between information representing the state of a predetermined space, information regarding flow regulators included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding flow within the predetermined space in that case, and can output information regarding flow within the predetermined space simply by inputting information representing the information of the predetermined space and information of the flow regulators. As a result, a user who has devised the arrangement of the flow regulators can recognize whether a desired result can be obtained regarding the flow in the predetermined space by the arrangement of the flow regulators devised by the user without performing a simulation.

Brief Description of the Drawings

[0018]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Best Mode for Carrying Out the Invention

[0019] <Embodiment> Hereinafter, an information processing apparatus according to the present invention will be described with reference to the drawings. First, a general description of the information processing apparatus will be given, and then specific usage examples of the information processing apparatus will be described.

[0020] <Configuration> The information processing apparatus 100 according to the present embodiment is a computer system that performs processing for estimating a result obtained based on input conditions. The information processing apparatus 100 is a device that inputs and outputs information related to a flow regulation body that regulates the flow of gas, liquid, heat, etc. within a predetermined space in a predetermined space. Further, the information processing apparatus 100 may generate a learning model necessary for executing the above-described estimation processing.

[0021] FIG. 1 is a block diagram showing a configuration example of the information processing apparatus 100. Here, the information processing apparatus 100 includes a communication unit 120, an input unit 130, a control unit 140, a storage unit 150, and an output unit 160. The communication unit 120, the input unit 130, the control unit 140, the storage unit 150, and the output unit 160 may be communicably connected to each other by a connection line 110.

[0022] The communication unit 120 is a communication interface that communicates with an external device via a network. The communication unit 120 transmits data received from an external device to the control unit 140. Further, the communication unit 120 transmits specified data to a specified external device or the like in accordance with an instruction from the control unit 140.

[0023] The input unit 130 is an input interface that receives input from the user of the information processing apparatus 100. The input unit 130 may be realized by, for example, a keyboard, a mouse, a touch panel, etc., but is not limited thereto. The input unit 130 may, for example, receive input by voice. When the input unit 130 receives input from the user, it transmits the received input content to the control unit 140.

[0024] The control unit 140 is a processor having a function for controlling each part of the information processing apparatus 100. The control unit 140 uses various programs and data stored in the storage unit 150 to execute the processing that the information processing apparatus 100 should execute, that is, the processing for estimating the flow of gas, liquid, solid, heat, etc. in a predetermined space.

[0025] In order to realize this processing, the control unit 140 includes a reception unit 141 and an estimation unit 142. Further, the control unit 140 may include a learning unit 143.

[0026] The reception unit 141 receives the information transmitted from the communication unit 120 or the input unit 130 as information to be input to the learning model 151. That is, the reception unit 141 receives the input of one-dimensional information indicating information representing the state of a predetermined space and information regarding the flow regulating bodies included in each small space when a plurality of small spaces included in the predetermined space are defined.

[0027] Here, the predetermined space will be described. The predetermined space is a space that the user wants to recognize the flow of things (for example, liquid, gas, solid, heat, etc.) in that space. The shape of the predetermined space is various, and it may have a more complex shape than a plane, a rectangular parallelepiped, a cylinder, a sphere, a hemisphere, etc. Also, the information representing the state of the predetermined space may be information such as the shape, volume, temperature when estimating the flow, the composition of the gas or liquid contained in the predetermined space, etc., and may be part or all of these, and further, information other than these may be included. It can also be said that the information representing the state of the predetermined space is information indicating the conditions for estimating the flow in the predetermined space.

[0028] Also, the plurality of small spaces included in a predetermined space may be spaces within the predetermined space and smaller than the predetermined space. Each of the plurality of small spaces may partially overlap. That is, among the plurality of small spaces, the small spaces may be defined such that a part of the first small space overlaps with a part of the second small space. Also, the plurality of small spaces do not necessarily have the same shape as each other, and may or may not be in a similar relationship with the predetermined space.

[0029] Figures 2(a) and (b) show examples of defining a plurality of small spaces within a predetermined space.

[0030] Figure 2(a) shows an example in which a rectangular parallelepiped-shaped predetermined space 200 is divided into rectangular parallelepiped shapes smaller than the predetermined space 200 to define small spaces.

[0031] Figure 2(b) shows another example of defining small spaces within a predetermined space, and shows an example in which a triangular prism-shaped space is divided to define small spaces.

[0032] Note that in Figures 2(a) and (b), an example of a rectangular parallelepiped-shaped space is shown as the predetermined space. However, as described above, the shape of the predetermined space is not limited to a rectangular parallelepiped shape, and may be a cylindrical shape, a hemispherical shape, or a more complex shape.

[0033] The flow regulator is for regulating the flow of a fluid flowing within a predetermined space, and may promote the regulated flow, change the flow, inhibit the flow, etc. For example, when the fluid flowing within the predetermined space is a liquid or a gas, it may be an inlet hole for allowing the liquid or gas to flow into the predetermined space, a discharge hole for discharging the liquid or gas within the predetermined space from the predetermined space, a plate or a propeller for changing the flow direction of the liquid or gas within the predetermined space, or a baffle plate for obstructing the flow of the liquid or gas within the predetermined space.

[0034] The estimation unit 142 inputs the information received by the reception unit 141 (information representing the state of a predetermined space and information regarding the flow regulating body) into the learning model 151 stored in the storage unit 150, and estimates the flow (flow of liquid, gas, solid, heat, etc., changes such as reaction of substances) in the predetermined space 200 formed according to the location where the flow regulating body is installed at a desired measurement point within the predetermined space 200.

[0035] FIG. 2(c) and (d) are diagrams showing examples of locations for estimating flow in a predetermined space.

[0036] FIG. 2(c) shows an example when a plane is set as the location 210 for estimating flow within the predetermined space 200.

[0037] Also, FIG. 2(d) shows an example when a space (a rectangular parallelepiped-shaped space) is set as the location 210 for estimating flow within the predetermined space 200.

[0038] In addition, the location for estimating flow may be anywhere within the predetermined space 200, and its range may be any. For example, it may be a point within the predetermined space 200, or it may be the surface of some object included within the predetermined space 200, but it is not limited thereto.

[0039] Note that in FIG. 2, an example where the predetermined space is a three-dimensional space is shown, but the predetermined space may be a plane, and the plurality of small spaces may be surfaces smaller than the plane included in that plane.

[0040] Returning to FIG. 1, the learning unit 143 receives, as teacher data, information representing the state of a predetermined space and information regarding fluid regulators included in each of a plurality of small spaces defined within the predetermined space, and information regarding the flow within the predetermined space in that case, from the communication unit 120 or the input unit 130. The learning unit 143 receives inputs of a plurality of pieces of teacher data and learns the relationship between the state of the predetermined space and the fluid regulators arranged in the plurality of small spaces within the predetermined space, and the flow within the predetermined space in that case. More specifically, the learning unit 143 learns the relationship between information representing the state of the predetermined space and one-dimensional information indicating the number of fluid regulators included in each of the plurality of small spaces defined within the predetermined space, and the flow within the predetermined space in that case, and generates a learning model 151. The learning unit 143 may store the generated learning model 151 in the storage unit 150.

[0041] Further, the learning unit 143 may receive inputs of the already stored learning model 151 and new information related to the learning model 151 and perform re-learning. The new information related to the learning model 151 is basically new teacher data associating an arrangement different from the arrangement of the learned fluid regulators with information regarding the flow within the predetermined space in that case.

[0042] The storage unit 150 is a storage medium that stores various programs and data required by the information processing apparatus 100. As an example, the storage unit 150 may be realized by an HDD (Hard Disc Drive), an SSD (Solid State Drive), a flash memory, or the like. Further, the storage unit 150 may be used as a ROM and a RAM as a work area when the control unit 140 executes processing.

[0043] The memory unit 150 stores the learning model 151. This learning model 151 may be a learning model generated by the learning unit 143 or may be generated by an external device. The learning model 151 is a learning model that has learned the relationship between information representing the state of a predetermined space and information regarding fluid regulators included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case.

[0044] The output unit 160 outputs information regarding the flow estimated by the control unit 140. The output of information by the output unit 160 may be transmitted to an external device via the communication unit 120, or, although not shown, may be output as an image or characters to a monitor installed or connected to the information processing apparatus 100, or may be output as sound from a speaker.

[0045] <Operation Example> FIG. 3 is a flowchart showing an example of the generation process of the learning model 151. Here, an example generated by the learning unit 143 is shown, but as described above, the learning model 151 may be generated by another device.

[0046] As shown in FIG. 3, the learning unit 143 receives input of information indicating the state of a predetermined space and information regarding fluid regulators included in each small space when a plurality of small spaces included in the predetermined space are defined, from the communication unit 120 or the input unit 130 (step S301).

[0047] Also, the learning unit 143 receives input of information regarding the flow within the predetermined space when the process is executed with the received state of the predetermined space and the arrangement of the fluid regulators in that case, from the communication unit 120 or the input unit 130 (step S302).

[0048] The learning unit 143 learns the relationship between (i) the information indicating the state of a predetermined space and the information regarding the flow regulators included in each small space when defining a plurality of small spaces included in the predetermined space, which is received in steps S301 and S302, and (ii) the information regarding the flow within the predetermined space (step S303). The learning unit 143 receives a plurality of pieces of teacher data that are pairs of the information received in steps S301 and S302, and performs learning. The algorithm used by the learning unit 143 for learning may be an existing algorithm for deep learning.

[0049] Based on the learning in step S304, the learning unit 143 generates a learning model 151. Then, the learning unit 143 stores the learning model 151 in the storage unit 150 (step S304) and ends the process. By this process, a learning model used for the estimation process is generated and stored in the storage unit 150 of the information processing apparatus 100.

[0050] FIG. 4 is a flowchart showing an operation example of the flow estimation process by the information processing apparatus 100.

[0051] The reception unit 141 of the control unit 140 receives, from the communication unit 120 or the input unit 130, the input of the information representing the state of a predetermined space and the information regarding the flow regulators included in each small space when defining a plurality of small spaces included in the predetermined space (step S401). The information received here is the information that the user wants to know about how the flow will be when processing is performed based on the input information and the state of the predetermined space and the arrangement of the flow regulators.

[0052] The estimation unit 142 inputs the information representing the state of the predetermined space and the information regarding the flow regulators, which are received by the reception unit 141, into the learning model 151 stored in the storage unit 150 (step S402). As described above, the learning model 151 is a model that has learned the relationship between the information indicating the state of a predetermined space and the information regarding the flow regulators included in each small space when defining a plurality of small spaces included in the predetermined space, and the information regarding the flow within the predetermined space.

[0053] As a result, the estimation unit 142 acquires information indicating the flow within a predetermined space in the case of the arrangement of the flow regulator received in step S401. Then, the estimation unit 142 causes the output unit 160 to output the acquired information (step S403), and ends the process.

[0054] Thereby, the user can obtain information regarding the flow within the predetermined space in that case, simply by inputting information representing the state of the predetermined space and information regarding the arrangement of the flow regulator devised by the user. Therefore, it is possible to specify whether the information regarding the arrangement of the input flow regulator is desirable for the user from the obtained information regarding the flow. Also, it is possible to back-calculate a more ideal arrangement from the obtained information regarding the flow.

[0055] Further, since a learning model that has learned the relationship between the information representing the state of the predetermined space and the information regarding the flow regulator included in each small space when a plurality of small spaces included in the predetermined space are defined, and the information regarding the flow within the predetermined space in that case is used, depending on the object of information processing, not only the arrangement of the flow regulator but also the state of the predetermined space can be optimized together. That is, as the learning model, by using a learning model that has learned the relationship between the state of the predetermined space and the information regarding the flow regulator included in each small space when a plurality of small spaces included in the predetermined space are defined, and the information regarding the flow within the predetermined space, depending on whether the obtained information regarding the flow is desirable as a result of the estimation process by the estimation unit 142, it can be used as a criterion for determining whether the state representing the input predetermined space state was suitable. Also, it is possible to back-calculate a more ideal state of the predetermined space from the obtained information regarding the flow.

[0056] Also, as shown in the above embodiment, by using, as information input to the learning model and also as part of the teacher data when generating the learning model, one-dimensional information regarding the flow regulating bodies included in each small space when a plurality of small spaces included in a predetermined space are defined, the estimation function of the information processing apparatus 100 can be improved. Details of the one-dimensional information are further described in Examples 1 to 5 below, but it can be said that this is information obtained by converting information regarding the arrangement of the flow regulating bodies included in a predetermined space.

[0057] As described above, information simply indicating the coordinates of each flow regulating body cannot directly represent the distribution of a plurality of flow regulating bodies, such as the relative positional relationship between each flow regulating body and between each flow regulating body and a specific point within a predetermined space. In contrast, the above one-dimensional information can represent the distribution of a plurality of flow regulating bodies by a combination of its elements, and can effectively represent the characteristics of the arrangement of the flow regulating bodies in a predetermined space. Therefore, according to the information processing apparatus 100 shown in the present embodiment, it becomes easier for the learning model to learn the relationship between information representing the state of a predetermined space and information regarding the arrangement of the flow regulating bodies included in the predetermined space, and information regarding the flow within the predetermined space in that case, and the estimation function acquired by learning can be improved. In short, in the information processing apparatus 100 according to the present embodiment, by using one-dimensional information as the object of learning rather than two-dimensional information such as the position coordinates of two-dimensional (or three-dimensional) flow regulating bodies, the content of learning for the information processing apparatus is simplified (simplified), enabling more accurate learning to be performed. As a result, the estimation result can be improved.

[0058] Hereinafter, the estimation by the information processing apparatus 100 as described above will be described using specific examples.

[0059] <Example 1> In Example 1, a deposition reactor where a deposition reaction is carried out will be taken as an example for explanation. That is, in the Berge apparatus, an example of a deposition reactor for depositing silicon on a silicon rod will be used for explanation. Here, an example of depositing silicon on a silicon rod is shown, but as the deposition reactor, any other deposition reaction may be used as long as a substance based on the injected gas is deposited on the core.

[0060] In the Berge apparatus, gas is introduced to deposit silicon on the silicon rod. In that case, it is preferable that silicon is deposited as uniformly as possible on the surfaces of all the silicon rods.

[0061] FIG. 5(a) is a perspective view schematically showing a reaction vessel 200a of the Berge apparatus as a deposition reactor. Further, FIG. 5(b) is a schematic side cross-sectional view of the reaction vessel 200a of the Berge apparatus. As shown in FIG. 5, in the reaction vessel 200a of the Berge apparatus, a gas nozzle 501 for injecting a gas related to the substance deposited in the reaction vessel is provided. In the example of FIG. 5, an example of arranging five gas nozzles 501 is shown, but the number of gas nozzles 501 is appropriate. Further, in the reaction vessel, a silicon rod which is a core on which a substance is deposited is erected. In the example of FIG. 5, an example of providing four silicon rods in the reaction vessel 200a is shown, but the number of silicon rods is also appropriate.

[0062] FIGS. 6 and 7 are diagrams showing an example of a small space included in a predetermined space in Example 1 and an arrangement example of a flow regulating body arranged therein. Using these figures, the input format of information input to the information processing apparatus 100 will be described in each case.

[0063] In FIGS. 6 and 7, the boundaries of the space are shown by dotted lines. The small circles in the figure indicate the arrangement positions of the gas nozzles 501. In FIGS. 6 and 7, for the sake of clarity of the drawing, reference numerals are given only to one gas nozzle.

[0064] FIG. 6(a) shows an example in which the space inside the Berger device 200a is regarded as a predetermined space and defined as a plurality of small spaces, i.e., space a1 and space a2 which are planes on the bottom surface of the Berger device 200a. In FIG. 6(a), the outer shape of the bottom surface of the reaction vessel 200a is divided into two spaces a1 and a2 in the radial direction with concentric circles to define small spaces. Also, FIG. 6(a) shows an example in which eight gas nozzles 501 are arranged. In the case of FIG. 6(a), no gas nozzle 501 is arranged in the small space a1, and eight gas nozzles 501 are arranged in the small space a2. When such small spaces and arrangements are made, the information regarding the flow regulator input to the learning model 151 during estimation is, as an example, one-dimensional information {a1,a2}={0,8} that is input. When simply regarding this bottom surface as a coordinate system without considering the normal small spaces, a plurality of two-dimensional information, i.e., the coordinates of each gas nozzle 501 in that coordinate system, will be input, so it can be said that the variables to be processed increase. Compared with such a case, in the aspect shown in this embodiment, input as one-dimensional information is possible, the arithmetic processing is easy, and the arithmetic time can also be shortened. Note that this one-dimensional information is also the information input during learning.

[0065] FIG. 6(b) shows an example in which the gas nozzles 501 are arranged in the same manner as in FIG. 6(a). On the other hand, FIG. 6(b) shows an example in which the definition of the small spaces is made not in the radial direction but in the circumferential direction, and four small spaces a3, a4, a5, and a6 are defined. In the example shown in FIG. 6(b), three gas nozzles 501 are arranged in the small space a3, one gas nozzle 501 is arranged in the small space a4, three gas nozzles 501 are arranged in the small space a5, and one gas nozzle 501 is arranged in the small space a6. In this case, the information regarding the flow regulator input to the learning model 151 during estimation can be expressed as, for example, one-dimensional information {a3,a4,a5,a6}={3,1,3,1} and can be input.

[0066] FIG. 7(a) is a case where the definition of small spaces is defined in the circumferential direction in the same manner as in FIG. 6(b). Although it is similar to the case of dividing into four small spaces, it shows an example where the method of defining the small spaces is different. In FIG. 6(b), it is divided in the diagonal direction of the drawing, whereas in FIG. 7(a), an example of dividing in the vertical and horizontal directions of the drawing is shown. As shown in FIG. 7(a), each of the small spaces a8, a9, a10, a11 shows a case where two gas nozzles 501 are arranged respectively. In this case, the information regarding the flow regulating body input to the learning model 151 during estimation can be expressed, for example, as one-dimensional information {a8, a9, a10, a11} = {2, 2, 2, 2} and can be input.

[0067] FIG. 7(b) shows an example of defining small spaces by dividing the bottom surface of the reaction vessel 200a in both the radial direction and the circumferential direction. In FIG. 7(b), an example of defining eight small spaces a13 to a20 is shown. And in the example of FIG. 7(b), the information regarding the flow regulating body input to the learning model 151 during estimation can be expressed, for example, as one-dimensional information {a13, a14, a15, a16, a17, a18, a19, a20} = {2, 0, 2, 0, 2, 2, 5, 2} and can be input.

[0068] Note that the examples of defining small spaces shown in FIGS. 6 and 7 are just examples, and each small space may be defined with different shapes and different ratios.

[0069] FIG. 8 is a flowchart showing an operation example related to the generation of a learning model by the information processing apparatus 100 to which the flowchart of FIG. 3 is applied in the case of Example 1.

[0070] As shown in FIG. 8, the learning unit 143 of the control unit 140 of the information processing apparatus 100 receives input of information such as the size, volume, and shape of the reaction vessel of the precipitation reactor, rod information regarding the arrangement of silicon rods, and nozzle information regarding the arrangement of gas nozzles (step S801).

[0071] The learning unit 143 receives information regarding deposition in the silicon rods when a deposition reaction is carried out under predetermined conditions based on the received information (step S802). Here, the information regarding deposition may be information such as the heat flux in the reaction vessel, the heat flux near each silicon rod, and the thickness of the silicon deposited on the surface of each silicon rod.

[0072] The learning unit 143 defines a surface smaller than the bottom surface included in the bottom surface of the deposition reactor and defines a plurality of small spaces (step S803). The definition of the small space here means determining the range of the small space in the deposition reactor as data.

[0073] The learning unit 143 counts the number of gas nozzles included in each small space based on the nozzle information received in step S801 and the small spaces defined in step S803 (step S804).

[0074] The learning unit 143 regards the number of gas nozzles included in each small space as information regarding the flow regulating body included in each small space, regards the information regarding deposition in the silicon rod as information regarding the flow in a predetermined space, and learns the relationship between the information representing the state of the predetermined space and the information regarding the flow regulating body and the information regarding the flow in the predetermined space (step S805). The learning unit 143 performs multiple learning on the relationship between the information representing the state of the predetermined space, the number of gas nozzles included in each small space, and the information regarding deposition in the silicon rod, and generates a learning model.

[0075] The learning unit 143 stores the generated learning model 151 in the storage unit 150 (step S806) and ends the process.

[0076] FIG. 9 is a flowchart showing an example of an estimation operation by the information processing apparatus 100 for obtaining information regarding deposition using the learning model 151 stored in the storage unit 150.

[0077] As shown in FIG. 9, the reception unit 141 of the control unit 140 receives, as information on the flow regulator, information indicating the state inside the precipitation reactor and arrangement information for arranging the gas nozzles with respect to the precipitation reactor (step S901). The reception unit 141 receives, for example, arrangement information indicating the arrangement of the gas nozzles input by the user to the input unit 130.

[0078] Based on the received arrangement information, the estimation unit 142 counts the number of gas nozzles included in a plurality of small spaces included in the bottom surface of the precipitation reactor (step S902). Note that the definition of the plurality of small spaces at this time is the same as the definition of the plurality of small spaces used when training the learning model 151. That is, for example, when creating the learning model 151, if the small spaces shown in FIGS. 6(a), (b), 7(a), and (b) are used, that is, even at the stage of estimation by the estimation unit 142, the small spaces a1 to a20 are used, and the number of gas nozzles included in each of the small spaces a1 to a20 is counted.

[0079] The estimation unit 142 inputs one-dimensional information in which information representing the state of a predetermined space and identification information for identifying the small space are associated with the number of gas nozzles included in the small space into the learning model 151 read from the storage unit 150 (step S903). By inputting the one-dimensional information into the learning model 151, information regarding the precipitation reaction can be obtained.

[0080] The estimation unit 142 instructs the output unit 160 to output information regarding the heat flux around the silicon rod as information regarding the precipitation reaction obtained from the learning model 151 (step S904), and ends the process.

[0081] Thereby, the user can recognize whether the input arrangement of the gas nozzles is an arrangement that can obtain a desired result without actually performing a simulation.

[0082] Note that, instead of the number of gas nozzles included in each small space, the information on the flow regulating bodies included in each small space received by the learning unit 143 may be the sum of the gas flow rates flowing into or out of the gas nozzles included in each small space, and it may be either of both (both the number of gas nozzles and the inflow amount, both the number of gas nozzles and the outflow amount, both the gas inflow amount and the gas outflow amount, or any of all of them). For example, taking the case of Fig. 6(b) as an example, assuming that the gas flow rates flowing from the gas nozzles located at the 12 o'clock position clockwise in order into a predetermined space from the gas nozzles are G1, G2, G3, G4, G5, G6, G7, G8. Then, the one-dimensional information based on the sum of the gas flow rates flowing into the gas nozzles can be expressed as {a3, a4, a5, a6} = {G1 + G2 + G8, G3, G4 + G5 + G6, G7}. Also, when using the number of gas nozzles and the gas inflow amount, assuming that the small spaces defining the number of gas nozzles corresponding to the small spaces a3, a4, a5, a6 in Fig. 6(b) are as1, as2, as3, as4, and the small spaces defining the gas inflow amounts corresponding to the small spaces a3, a4, a5, a6 in Fig. 6(b) are ag1, ag2, ag3, ag4, the one-dimensional information can be expressed as {as1, as2, as3, as4, ag1, ag2, ag3, ag4} = {3, 1, 3, 1, G1 + G2 + G8, G3, G4 + G5 + G6, G7}.

[0083] Note that in Fig. 9, the estimation unit 142 is configured to count the number of gas nozzles included in each small space. However, in step S901, it may be configured to receive the information indicating the small space and the information indicating the number of gas nozzles included in that small space. In that case, the process of step S902 can be omitted. That is, the one-dimensional input information described with reference to Figs. 6 and 7 may be input via the communication unit 120 or the input unit 130.

[0084] According to the first embodiment, by inputting the information representing the state inside the deposition reactor, the arrangement of the gas nozzles, or the number of gas nozzles included in the small spaces, information regarding the deposition reaction can be obtained.

[0085] Note that, for Example 1, the inventors created a learning model by inputting a plurality of randomly created information (information representing the state of a randomly created predetermined space, information regarding the arrangement of gas nozzles in a small space, and information regarding the convective heat flux on the surface of the silicon rod at that time), and evaluated the estimation results obtained by inputting different random information (information representing the state of a randomly created predetermined space and information regarding the arrangement of gas nozzles in a small space) into the created learning model. That is, when comparing the result of an experiment conducted under the same conditions as the information input to the learning model with the estimation result of the learning model, the coefficient of determination of the estimation result by the learning model could be made 0.7 or more. That is, it was possible to obtain a result that the accuracy (correctness) of the estimation result by the learning model was 70% or more. Regarding this coefficient of determination, further improvement can be expected by increasing the number of data to be learned (the number of teacher data) or by selecting the teacher data to be learned. Note that although examples of randomly created information are described above, of course, similar results can also be obtained by learning information that is not random but specified in advance in the same manner.

[0086] On the other hand, the inventors also created and evaluated a learning model when dealing with xy coordinates instead of the number of gas nozzles arranged with respect to a small space. As a result, the coefficient of determination was less than 0.4. That is, it can be said that the prediction performance of the flow is greatly improved by using a learning model defined by the number of flow regulating bodies arranged in a small space rather than using a learning model that defines the flow regulating bodies in a predetermined space by coordinates. Therefore, in the case where flow regulating bodies are arranged in a predetermined space, in the aspect of predicting the flow, an improvement in the estimation result can be expected rather than using the position coordinates of the flow regulating bodies in the predetermined space, and this is the same in Examples 2 to 5 shown hereinafter.

[0087] <Example 2> In the above Example 1, an estimation example in a precipitation reactor is shown. In this Example 2, an example of finely dividing the gas bubbles as much as possible when mixing gas into the powder in a fluidized bed reactor will be described.

[0088] FIG. 10 is a diagram schematically showing a fluidized bed reactor. The fluidized bed reactor 200b is filled with powder and granules, and gas is mixed therein. It is desirable that the gas bubbles are mixed in as small and finely divided as possible. However, as the bubbles rise inside the apparatus, they coalesce and grow larger, so it is necessary to re-divide them with a baffle plate. The bubble 1001 is divided by colliding with the baffle plate 1002. For this purpose, it is desirable to provide the baffle plate 1002 so that the gas flowing in from the gas inlet (see the arrow in the drawing) collides with the fluidized bed reactor as much as possible and is pulverized so that the size of the bubbles becomes as uniform as possible. As shown in FIG. 10(a), as the bubble 1001 moves toward the upper layer of the drawing of the fluidized bed reactor, it coalesces and grows larger. However, thereafter, as shown in FIG. 10(b), the bubble 1001 that has collided with the baffle plate 1002 becomes small again.

[0089] In Example 2, the information processing apparatus 100 estimates the size or particle size of the bubbles based on the arrangement of the baffle plate 1002, which is a flow regulating body.

[0090] FIG. 11 shows an example of the definition of a small space in Example 2. FIG. 11(a) shows an example in which the fluidized bed reactor 200b is defined in the horizontal direction as four small spaces b1, b2, b3, and b4. When the small space is defined as shown in FIG. 11(a), three baffle plates are arranged in the small space b2 and two baffle plates are arranged in the small space b3. Therefore, in this case, the information regarding the flow regulating body input to the learning model 151 during estimation can be expressed, for example, as one-dimensional information {b1, b2, b3, b4} = {0, 3, 2, 0} and can be input.

[0091] Further, FIG. 11(b) shows an example in which the fluidized bed reactor 200b is defined in the horizontal direction into five small spaces b1, b2, b3, b4, b5. When the small spaces are defined as shown in FIG. 11(b), five baffle plates are arranged in the small space b2 and three baffle plates are arranged in the small space b4. Therefore, in this case, the information on the fluid regulating body input to the learning model 151 during estimation can be expressed, for example, as one-dimensional information {b5, b6, b7, b8, b9} = {0, 5, 0, 3, 0} and can be input.

[0092] FIG. 12 is a flowchart showing an operation example related to the generation of a learning model by the information processing apparatus 100 to which the flowchart of FIG. 3 is applied in the case of Example 2.

[0093] As shown in FIG. 12, the learning unit 143 of the control unit 140 of the information processing apparatus 100 receives input of information such as the size, volume, and shape of the fluidized bed reactor, information on the arrangement position of the gas inlet, and arrangement information regarding the arrangement of the baffle plates (step S1201).

[0094] Based on the received information, the learning unit 143 receives information regarding bubble changes in the fluidized bed reactor when gas is allowed to flow in from the gas inlet under predetermined conditions (step S1202). Here, the information regarding bubble changes may be information such as the size and particle size of bubbles at a desired measurement position.

[0095] The learning unit 143 defines a plurality of small spaces inside the fluidized bed reactor (step S1203). The definition of the small space here means determining the range of the small space in the fluidized bed reactor as data, and in the example of FIG. 11, it refers to each region separated by the plane indicated by the dotted line in the fluidized bed reactor 200b. When the change of bubbles is important as in the second embodiment, it is effective to define the small space so as to be separated in the flow direction of the bubbles.

[0096] In step S1204, the learning unit 143 counts the number of baffle plates included in each small space based on the information representing the state in the fluidized bed reactor received in step S1201, the arrangement information of the baffle plates, and the small spaces defined in step S1203.

[0097] The learning unit 143 regards the number of baffle plates included in each small space as information regarding the fluid regulating bodies included in each small space, regards the information regarding the state of the bubbles at each position in the fluidized bed reactor as information regarding the flow in a predetermined space, and learns the relationship between the information representing the state in the fluidized bed reactor, the information regarding the fluid regulating bodies, and the information regarding the flow (step S1205). The learning unit 143 performs multiple learning on the relationship between the information representing the state in the fluidized bed reactor, the number of baffle plates included in each small space, and the information regarding the state of the bubbles in the fluidized bed reactor, and generates a learning model.

[0098] The learning unit 143 stores the generated learning model 151 in the storage unit 150 (step S1206) and ends the process.

[0099] FIG. 13 is a flowchart showing an example of the estimation operation by the information processing apparatus 100 for obtaining information regarding the state of the bubbles using the learning model 151 stored in the storage unit 150.

[0100] As shown in FIG. 13, the reception unit 141 of the control unit 140 receives the arrangement information for arranging the baffle plates with respect to the fluidized bed reactor as the information of the fluid regulating bodies (step S1301). The reception unit 141 receives, for example, the arrangement information indicating the arrangement of the baffle plates input by the user to the input unit 130.

[0101] Based on the received arrangement information, the estimation unit 142 counts the number of baffle plates included in a plurality of small spaces (step S1302). Note that the definition of the plurality of small spaces at this time is the same as the definition of the plurality of small spaces used when learning the learning model 151.

[0102] The estimation unit 142 inputs one-dimensional information associating information representing the state in the fluidized bed reactor and identification information for identifying the small space with the number of baffle plates included in the small space into the learning model 151 read from the storage unit 150 (step S1303). By inputting the one-dimensional information into the learning model 151, information regarding the state of bubbles at a desired location in the fluidized bed reactor can be obtained.

[0103] The estimation unit 142 instructs the output unit 160 to output the information regarding the state of bubbles obtained from the learning model 151 (step S1304), and ends the process.

[0104] As shown in the second embodiment, the user can recognize whether the bubbles have a desired size and particle size with the arrangement of the baffle plates arranged by the user in the fluidized bed reactor. Further, from the obtained results, it is possible to estimate the arrangement of the baffle plates likely to make the bubbles in a desired state by inverse calculation.

[0105] Note that the information regarding the flow regulating bodies included in each small space received by the learning unit 143 may be, in addition to the number of baffle plates included in each small space, the sum of the areas of the baffle plates included in each small space, or both. That is, as an example, in the case where the baffle plates are arranged as shown in FIG. 11(a), when the areas of the baffle plates arranged in the small space b2 are A1, A2, and A3 respectively, and the areas of the baffle plates arranged in the small space b3 are A4 and A5 respectively, the teacher data of the learning model or the information input to the learning model when using the sum of the areas of the baffle plates can be expressed as {b1, b2, b3, b4} = {0, A1 + A2 + A3, A4 + A5, 0}. Further, in the case of using both, when the small space defined by the number of baffle plates is ba and the small space defined by the area of the baffle plates is bb, it can be expressed as {ba1, ba2, ba3, ba4, bb1, bb2, bb3, bb4} = {0, 3, 2, 0, 0, A1 + A2 + A3, A4 + A5, 0}. Here, ba1 and bb1 are the same space as the small space b1 in FIG. 11(a) as the small space, and the same applies to the other small spaces.

[0106] In the present embodiment, information for outputting (evaluating) the size (diameter) and particle size of bubbles is used, but this may be the number of bubbles included within a desired measurement range. For example, it can be recognized that when the number of bubbles is large, the diameter of the bubbles included within the desired measurement range is small.

[0107] <Example 3> In Example 3, an example of estimating the temperature at a specific location within a desired space will be described.

[0108] FIG. 14 is a diagram showing an example of a desired space according to Example 3. FIG. 14 is a plan view showing an example including a heating element 1401 and ventilation openings 1402a and 1402b indoors as an example. As the heating element 1401, for example, it may be a device that becomes high temperature by some process, or it may be some heater or the like. Since there are some heating elements 1401 that desirably maintain a certain temperature range for convenience of processing, it is desirable to know in advance the conditions for maintaining the desired temperature range. In the figure, the ventilation opening 1402a is an opening for taking in external air, and the ventilation opening 1402b is a ventilation fan for discharging indoor air to the outside.

[0109] In Example 3, an estimation will be described for the case of temperature control by providing ventilation openings or the like and performing ventilation in an object with a temperature change such as the heating element 1401.

[0110] FIG. 15 is a perspective view schematically showing a desired space according to Example 3, in which the heating element 1401 is provided. Also, FIG. 15 is a diagram for explaining an example of defining a plurality of small spaces in Example 3.

[0111] As shown in FIG. 15, in room 200c, a heating element 1401 is provided indoors. Ventilation is performed so that the temperature of this heating element 1401 does not become higher than a predetermined value. For this purpose, in the example shown in FIG. 15, four ventilation openings 1402a are provided in the wall surface 1512 on the back side of the drawing of room 200c, and two ventilation openings 1402b are provided in the wall surface 1511 on the front side of the drawing. As an example, the wall surface 1512 is divided into four regions by the dotted line shown in the figure, and are respectively defined as small spaces a1, a2, a3, and a4. Also, as an example, the wall surface 1511 is divided into four regions by the dotted line shown in the figure, and are respectively defined as small spaces a5, a6, a7, and a8. Since these ventilation openings 1402a and 1402b are involved in the air flow in room 200c and are involved in the up and down movement of the temperature of the heating element 1401, they can be defined as flow regulating bodies.

[0112] Therefore, in the example of FIG. 15, the information on the number of flow regulating bodies included in the small spaces input to the learning model 151 can be expressed as one-dimensional information {c1, c2, c3, c4, c5, c6, c7, c8} = {1, 1, 1, 1, 1, 1, 0, 0}. As shown in FIG. 15, a predetermined space may be defined as one space by grouping together locations that are separated from each other. Also, when the ventilation opening 1402a is an air supply port (in side (in) for taking in air) and the ventilation opening 1402b is an exhaust port (out side (out) for exhausting air), their respective arrangements can be treated as separate information. Therefore, the small spaces input to the learning model 151 and the information on the number of flow regulating bodies included in the small spaces can be expressed as one-dimensional information {c1_in, c2_in, c3_in, c4_in, c5_in, c6_in, c7_in, c8_in, c1_out, c2_out, c3_out, c4_out, c5_out, c6_out, c7_out, c8_out} = {1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0}.

[0113] FIG. 16 is a flowchart showing an operation example related to the generation of a learning model by an information processing apparatus 100 to which the flowchart of FIG. 3 is applied in the case of Example 3.

[0114] As shown in FIG. 16, the learning unit 143 of the control unit 140 of the information processing apparatus 100 receives input of information such as the size, volume, and shape of the room in which the heating element 1401 is provided, information on the position of the heating element 1401, and arrangement information regarding the arrangement of ventilation openings (step S1601). Information such as the position information of the heating element 1401 may also be received.

[0115] Based on the received information, the learning unit 143 receives information regarding the temperature change in the room when air flows in from the ventilation opening 1402a under predetermined conditions (step S1602). Here, the information regarding the temperature change is the temperature information at a desired measurement position, and may be information regarding the temperature rise or temperature drop per unit time.

[0116] The learning unit 143 creates a plurality of regions on the side wall of the room and defines a plurality of small spaces (step S1603). The definition of the small space here means determining the range of the small space in the room as data. In the example of FIG. 15, it refers to each of the regions c1 to c8 separated by the plane indicated by the dotted lines on the wall surfaces 1511 and 1512 of the room 200c.

[0117] Based on the information regarding the predetermined space received in step S1601, the ventilation opening arrangement information, and the small spaces defined in step S1603, the learning unit 143 counts the number of ventilation openings included in each small space (step S1604). The information regarding the predetermined space here is the information regarding the interior 200c, and may include the volume and shape of the interior, as well as the heat generation amount of the heating element 1401.

[0118] The learning unit 143 regards the number of ventilation openings included in each small space as information regarding the flow regulation body included in each small space, regards the information regarding the temperature change in the room as information regarding the flow in the predetermined space, and learns the relationship between the information regarding the predetermined space, the information regarding the flow regulation body, and the information regarding the flow (step S1605). The learning unit 143 performs multiple learning on the relationship between the information regarding the predetermined space, the number of ventilation openings included in each small space, and the information regarding the temperature change in the room, and generates a learning model.

[0119] The learning unit 143 stores the generated learning model 151 in the storage unit 150 (step S1606) and ends the process.

[0120] FIG. 17 is a flowchart showing an example of an estimation operation by the information processing apparatus 100 for obtaining information regarding a temperature change at a desired location using the learning model 151 stored in the storage unit 150.

[0121] As shown in FIG. 17, the reception unit 141 of the control unit 140 receives information regarding the room 200c and arrangement information for arranging ventilation openings for the room as information on the flow regulation body (step S1701). The reception unit 141 receives, for example, arrangement information indicating the arrangement of the ventilation openings input from the user to the input unit 130.

[0122] The estimation unit 142 counts the number of ventilation openings included in a plurality of small spaces based on the received arrangement information (step S1702). Note that the definition of the plurality of small spaces at this time is the same as the plurality of small spaces used when learning the learning model 151.

[0123] The estimation unit 142 inputs one-dimensional information in which information regarding a predetermined space and identification information for identifying a small space are associated with the number of ventilation openings included in each small space into the learning model 151 read from the storage unit 150 (step S1703). By inputting the one-dimensional information into the learning model 151, information regarding a temperature change at a desired location in the room can be obtained.

[0124] The estimation unit 142 instructs the output unit 160 to output the information regarding the temperature change obtained from the learning model 151 (step S1704) and ends the process. The information regarding the temperature change may be the temperature of the room that is finally reached, or may be the amount of temperature change or the like.

[0125] As shown in Example 3, in a room, the user can estimate whether a desired temperature can be achieved at a certain location in the room, for example, at the heat source 1401, based on the arrangement of the ventilation openings he or she has arranged. Also, from the obtained results, it is possible to estimate the arrangement of the ventilation openings that would likely bring the indoor temperature to the desired temperature by back-calculation.

[0126] Note that the information regarding the flow regulating bodies included in each small space received by the learning unit 143 may be, in addition to the number of ventilation openings included in each small space, the sum of the opening areas of the ventilation openings included in each small space, or both. Also, instead of the area, the amount of gas discharged by each ventilation opening per unit time may be used, or the number of ventilation openings and the sum of the areas may be combined. For example, in the case where the opening areas of the ventilation openings provided in the small spaces c1, c2, c3, c4, c5, c6 in FIG. 15 are CA1, CA2, CA3, CA4, CA5, CA6 respectively, {c1, c2, c3, c4, c5, c6, c7, c8} = {CA1, CA2, CA3, CA4, CA5, CA6, 0, 0} can be used as the one-dimensional information when using the sum of the opening areas. Also, when defining the small spaces corresponding to the number of ventilation openings for the small spaces c1, c2, c3, c4, c5, c6, c7, c8 in FIG. 6 as ca1, ca2, ca3, ca4, ca5, ca6, ca7, ca8, and defining the small spaces for defining the sum of the opening areas as cb1, cb2, cb3, cb4, cb5, cb6, cb7, cb8, the one-dimensional information can be expressed as {ca1, ca2, ca3, ca4, ca5, ca6, ca7, ca8, cb1, cb2, cb3, cb4, cb5, cb6, cb7, cb8} = {1, 1, 1, 1, 1, 1, 0, 0, CA1, CA2, CA3, CA4, CA5, CA6, 0, 0}.

[0127] Here, although the temperature change of the heating element 1401 is estimated, this may be the surface temperature of the heating element 1401, the internal temperature, or even the temperature at a specific location of the heating element 1401. Alternatively, it may estimate the temperature at a specific location in the room. That is, due to the presence of the heating element 1401, in some cases, the room may become hot, which may cause inconvenience to people passing through it. Therefore, the temperature change at points on the path where people in the room pass may be estimated. Also, in Example 3, the case of air cooling was shown, but it is also possible to fill Room 200c with water or the like and perform temperature adjustment by allowing water to flow in and out through openings 1402a and 1402b.

[0128] <Example 4> In this Example 4, an example of estimating a temperature different from that in Example 3 will be described. In the above Example 3, an example of temperature adjustment (cooling) by a ventilation opening was described, but in this Example 4, the estimation of the temperature change in the case of directly cooling by a cooling tube will be described.

[0129] FIG. 18 is a plan view schematically showing a cooling device 200d. As shown in FIG. 18, a cooling tube 1801 is arranged to keep the inside of the cooling device 200d below a predetermined temperature, and the inside is filled with a coolant. As a result, the inside of the cooling device 200d is cooled. At this time, depending on the arrangement of the cooling tube 1801, the way of cooling inside the cooling device 200d is different (it may be cooled faster or slower depending on the location). Therefore, the cooling tube 1801 can be defined as a flow regulating body for regulating the temperature flow inside the cooling device 200d.

[0130] Therefore, in this Example 4, with the inside of the cooling device 200d as a predetermined space, it is estimated how the cooling tube 1801 is arranged and how it is cooled at a desired measurement point (for example, measurement point 1803).

[0131] Further, FIG. 18 shows an example of the definition of small spaces in Example 4. FIG. 18(a) shows an example in which when the cooling device 200d is viewed in a plane, it is defined into eight small spaces d1, d2, d3, d4, d5, d6, d7, d8. When the small spaces are defined as shown in FIG. 18, in the example of FIG. 18, one cooling tube is arranged across the small spaces a1, a3, a5, and one cooling tube is arranged across the small spaces d4, d6, d8. Therefore, in this case, the information regarding the fluid regulator input to the learning model 151 during estimation can be expressed, as an example, by one-dimensional information {d1, d2, d3, d4, d5, d6, d7, d8} = {1, 0, 1, 1, 1, 1, 0, 1} and can be input. Here, although one cooling tube is counted as one if it is included in one small space, a value obtained by quantifying the ratio of the cooling tube occupying the small space may be used as the number of fluid regulators.

[0132] FIG. 19 is a flowchart showing an operation example related to the generation of a learning model by the information processing apparatus 100 to which the flowchart of FIG. 3 is applied in the case of Example 4.

[0133] As shown in FIG. 19, the learning unit 143 of the control unit 140 of the information processing apparatus 100 receives input of information such as the size, volume, and shape of the cooling device, information on the arrangement position of the gas inlet, and arrangement information regarding the arrangement of the cooling tubes (step S1901).

[0134] Based on the received information, the learning unit 143 receives information regarding the temperature change in the cooling device when the cooling tube is filled with a coolant under predetermined conditions (step S1902). Here, the information regarding the temperature change may be information such as the degree of increase or decrease in temperature and the temperature after change at a desired measurement position.

[0135] The learning unit 143 creates a plurality of regions based on the top surface of the cooling device to define a plurality of small spaces (step S1903). The definition of the small space here means determining the range of the small space in the cooling device as data. In the example of FIG. 18, the top surface of the cooling device 200d refers to each region separated by a dotted line.

[0136] Based on the arrangement information of the cooling tubes received in step S1901 and the small spaces defined in step S1903, the learning unit 143 counts the number of cooling tubes included in each small space (step S1904).

[0137] The learning unit 143 regards the number of cooling tubes included in each small space as information regarding the flow regulating body included in each small space, regards the information regarding the temperature change at each position in the cooling device as information regarding the flow in a predetermined space, and learns the relationship between the information regarding the cooling device and the information regarding the flow regulating body and the information regarding the flow (step S1905). The learning unit 143 performs multiple learning on the relationship between the information regarding the cooling device and the number of cooling tubes included in each small space and the information regarding the temperature change in the cooling device, and generates a learning model.

[0138] The learning unit 143 stores the generated learning model 151 in the storage unit 150 (step S1906) and ends the process.

[0139] FIG. 20 is a flowchart showing an example of the estimation operation by the information processing apparatus 100 for obtaining information regarding the temperature change in the cooling device 200d using the learning model 151 stored in the storage unit 150.

[0140] As shown in FIG. 20, the reception unit 141 of the control unit 140 receives arrangement information indicating the position where the cooling tubes are to be arranged with respect to the cooling device as information regarding the flow regulating body (step S2001). The reception unit 141 receives, for example, arrangement information indicating the arrangement of the cooling tubes input by the user to the input unit 130.

[0141] Based on the received arrangement information, the estimation unit 142 counts the number of cooling tubes included in a plurality of small spaces (step S2002). At this time, the definition of the plurality of small spaces is the same as the definition of the plurality of small spaces used when training the learning model 151.

[0142] The estimation unit 142 inputs one-dimensional information associating information about a predetermined space and identification information for identifying a small space with the number of cooling tubes included in the small space into the learning model 151 read from the storage unit 150 (step S2003). By inputting the one-dimensional information into the learning model 151, information regarding temperature changes at a desired location within the cooling device can be obtained.

[0143] The estimation unit 142 instructs the output unit 160 to output the information regarding the temperature change obtained from the learning model 151 (step S2004), and ends the process.

[0144] As shown in the fourth embodiment, the user can recognize whether the inside of the cooling device reaches a desired temperature or shows a temperature change with the arrangement of the cooling tubes arranged by the user in the cooling device. Further, based on the obtained result, it is possible to estimate the arrangement of the cooling tubes that is likely to bring the temperature to a desired state.

[0145] Note that the information about the flow regulating bodies contained in each small space received by the learning unit 143 may be, in addition to the number of cooling tubes contained in each small space, the sum of the heat removal amounts per unit time of the cooling tubes contained in each small space, or both. For example, when the heat removal amounts (which may be, for example, the reduced temperature or the changed heat amount) by which the inside of the small spaces are cooled within a unit time by the cooling tubes contained in the small spaces d1, d3, d4, d5, d6, d8 in FIG. 18 are defined as DA1, DA3, DA4, DA5, DA6, DA8, respectively, the one-dimensional information can be expressed as {d1, d2, d3, d4, d5, d6, d7, d8} = {DA1, 0, DA3, DA4, DA5, DA6, 0, DA8}. Also, when using both the number of cooling tubes and the heat removal amount, if the small spaces corresponding to the number of cooling tubes are defined as da1, da2, da3, da4, da5, da6, da7, da8 and the small spaces corresponding to the heat removal amounts of the cooling tubes are defined as db1, db2, db3, db4, db5, db6, db7, db8, the one-dimensional information can be expressed as {da1, da2, da3, da4, da5, da6, da7, da8, db1, db2, db3, db4, db5, db6, db7, db8} = {1, 0, 1, 1, 1, 1, 0, 1, DA1, 0, DA3, DA4, DA5, DA6, 0, DA8}.

[0146] <Example 5> In this Example 5, it is an example of estimating the reaction in the same container as in Example 1, and an example different from Example 1 will be described.

[0147] FIG. 21 is a schematic perspective view showing an example of a reaction apparatus 200e for reacting a gas with a substance. The reaction apparatus 200e may be a precipitation reaction apparatus. For example, a gas is inserted into the reaction apparatus 200e through a plurality of gas inlet holes 2102 provided on the upper surface 2101 of the reaction apparatus 200e. Then, it reacts with the object 2104. Then, the gas in the reaction apparatus 200e is appropriately discharged from the discharge hole 2103 provided on the lower surface 2105. In such a mode, since the gas inlet hole 2102 and the discharge hole 2103 determine the flow in the reaction apparatus 200e, that is, the gas flow or the reaction in the object 2104, they can be defined as flow regulating bodies.

[0148] FIG. 22 shows an example of the definition of small spaces in Example 5. FIG. 22(a) defines six small spaces e1, e2, e3, e4, e5, e6 on the upper surface 2101 of the reaction apparatus 200e, and FIG. 22(b) shows an example in which six small spaces e7, e8, e9, e10, e11, e12 are defined on the lower surface 2012. When the small spaces are defined as shown in FIG. 22(a), one gas inlet hole 2102 is arranged in each of the small spaces e1 to e6. Therefore, in this case, the information regarding the flow regulating body input to the learning model 151 during estimation can be expressed as one-dimensional information such as {e1, e2, e3, e4, e5, e6} = {1, 2, 1, 1, 1, 1} as an example and can be input.

[0149] Also, as shown in FIG. 22(b), when the discharge hole 2103 is provided, the information regarding the flow regulating body input to the learning model 151 during estimation can be expressed as one-dimensional information such as {e7, e8, e9, e10, e11, e12} = {0, 1, 0, 0, 1, 0} as an example and can be input. Since the gas inlet hole 2102 (in) and the discharge hole 2103 (out) can be separated as separate information, in this case, it can also be expressed as 24-digit one-dimensional information in which in and out are separated as in Example 3.

[0150] FIG. 23 is a flowchart showing an operation example related to the generation of a learning model by the information processing apparatus 100 to which the flowchart of FIG. 3 is applied in the case of Example 5.

[0151] As shown in FIG. 23, the learning unit 143 of the control unit 140 of the information processing apparatus 100 receives input of information such as the size, volume, and shape of the reaction vessel, information on the arrangement position of the gas inlet, and arrangement information regarding the arrangement of the gas inlet holes (discharge holes) (step S2301).

[0152] Based on the received information, the learning unit 143 receives information regarding the reaction in the object 2104 in the reaction vessel when gas is introduced from the gas inlet under predetermined conditions (step S2302). Here, the information regarding the reaction may be information such as whether a desired reaction is occurring in the object 2104, whether the reaction is occurring evenly, etc. Further, when the reaction is a precipitation reaction, it may be information on the amount (thickness) of precipitation. Also, when a specific substance is generated (or consumed) by the reaction, it may be the concentration within a predetermined range of space (measurement point) of the specific substance.

[0153] The learning unit 143 creates a plurality of regions based on the side surface of the reaction vessel and defines a plurality of small spaces (step S2303). The definition of the small space here means determining the range of the small space in the reaction vessel as data, and in the example of FIG. 22, it refers to each region separated by the plane indicated by the dotted line of the reaction apparatus 200e.

[0154] Based on the arrangement information of the gas inlet holes (discharge holes) received in step S2301 and the small spaces defined in step S2303, the learning unit 143 counts the number of gas inlet holes (discharge holes) included in each small space (step S2304).

[0155] The learning unit 143 learns, as information regarding the flow regulating bodies included in each small space, the number of gas inlet holes (discharge holes) included in each small space, and learns, as information regarding the flow within a predetermined space, information regarding the reaction in the object 2104 in the reaction vessel (step S2305). The learning unit 143 performs multiple learning on the relationship between the number of gas inlet holes (discharge holes) included in each small space and the information regarding the reaction in the reaction vessel, and generates a learning model.

[0156] The learning unit 143 stores the generated learning model 151 in the storage unit 150 (step S2306), and ends the process.

[0157] FIG. 24 is a flowchart showing an example of the estimation operation by the information processing apparatus 100 for obtaining information regarding the reaction in the object using the learning model 151 stored in the storage unit 150.

[0158] As shown in FIG. 24, the reception unit 141 of the control unit 140 receives, as information regarding the flow regulating body, arrangement information for arranging gas inlet holes (discharge holes) with respect to the reaction vessel (step S2401). The reception unit 141 receives, for example, arrangement information indicating the arrangement of the gas inlet holes (discharge holes) input from the user to the input unit 130.

[0159] Based on the received arrangement information, the estimation unit 142 counts the number of gas inlet holes (discharge holes) included in a plurality of small spaces (step S2402). Note that the definition of the plurality of small spaces at this time is the same as the plurality of small spaces used when learning the learning model 151.

[0160] The estimation unit 142 inputs, into the learning model 151 read from the storage unit 150, one-dimensional information associating information regarding a predetermined space and identification information for identifying the small space with the number of gas inlet holes (discharge holes) included in the small space (step S2403). By inputting the one-dimensional information into the learning model 151, information regarding the reaction in the object at a desired location in the reaction vessel can be obtained.

[0161] The estimation unit 142 instructs the output unit 160 to output information regarding the reaction in the object obtained from the learning model 151 (step S2404), and ends the process.

[0162] As shown in Example 5, the user can recognize whether a desired reaction occurs in the reaction vessel with the arrangement of the gas inlet holes (discharge holes) that the user has arranged. Further, from the obtained results, it is possible to estimate the arrangement of the gas inlet holes (discharge holes) in which the reaction in the object is likely to be in a desired state by inverse calculation.

[0163] Note that the information regarding the flow regulators included in each small space, which the learning unit 143 receives, may be the sum of the amounts of gas flowing in (out) from the gas inlet holes (outlet holes) included in each small space, in addition to the number of gas inlet holes (outlet holes) included in each small space, or both. Here, "both" may mean defining each of the gas inflow amount and the discharge amount, or defining each of the number of gas inlet holes (outlet holes) and the gas inflow amount (discharge amount), or defining all of these. That is, instead of the number of gas inlet holes (outlet holes) shown in the above Example 5, the total sum of the amounts of gas flowing in per unit time (or a predetermined time period) from the gas inlet holes (outlet holes) included in each respective small space may be used. As an example, taking FIG. 22 as an example, for instance, if the amount of gas flowing in per unit time from the gas inlet holes included in the small space e1 is E1 liters, and from one of the gas inlet holes included in the small space e2, E2 liters of gas flows in per unit time, and from the other gas inlet hole, E3 liters of gas flows in per unit time. Then, in this case, the one-dimensional information input to the learning model is {e1, e2,...} = {E1, E2 + E3,...}. Thus, it is also possible to define and perform estimation of the learning model by the gas inflow amount (discharge amount) instead of the number of gas inlet holes (outlet holes). When defining the learning model and the one-dimensional information to be input by the number of gas inlet holes (outlet holes), it can be said that it is effective when the amount of gas flowing in (out) from each gas inlet hole (outlet hole) is constant for any hole. Also, when defining the learning model and the one-dimensional information to be input by the amount of gas flowing in (out), it can be said that it is effective when the amount of gas flowing in (out) from each gas inlet hole (outlet hole) is not constant.

[0164] Note that the programs of the embodiments of the present disclosure may be provided in a state stored in a computer-readable storage medium. The storage medium can store the program in a "non-transitory tangible medium". The storage medium can include any suitable storage medium such as HDDs and SSDs, or any suitable combination of two or more of these. The storage medium may be volatile, non-volatile, or a combination of volatile and non-volatile. Note that the storage medium is not limited to these examples, and any device or medium can be used as long as it can store the program.

[0165] Note that the information processing apparatus 100 can realize the functions of the plurality of functional units shown in the embodiments by, for example, reading out the program stored in the storage medium and executing the read program. Also, the program may be provided to the information processing apparatus 100 via any transmission medium (such as a communication network or a broadcast wave). The information processing apparatus 100 realizes the functions of the plurality of functional units shown in the embodiments by executing, for example, a program downloaded via the Internet or the like.

[0166] Note that the program can be implemented using, for example, script languages such as ActionScript and JavaScript (registered trademark), object-oriented programming languages such as Objective-C and Java (registered trademark), and markup languages such as HTML5.

[0167] At least a part of the processing in the information processing apparatus 100 may be realized by cloud computing configured by one or more computers. Also, each functional unit of the information processing apparatus 100 may be realized by one or a plurality of circuits that realize the functions shown in the above embodiments, and the functions of a plurality of functional units may be realized by one circuit.

[0168] Also, although the embodiments of the present disclosure have been described based on the drawings and examples, it should be noted that those skilled in the art can easily make various modifications and corrections based on the present disclosure. Therefore, it should be noted that these modifications and corrections are included in the scope of the present disclosure. For example, the functions included in each means, each step, etc. can be rearranged so as not to be logically contradictory, and a plurality of means, steps, etc. can be combined into one or divided. Also, the configurations shown in each embodiment may be appropriately combined.

Explanation of Reference Numerals

[0169] 100 Information processing apparatus 110 Connection line 120 Communication unit 130 Input unit 140 Control unit 141 Reception unit 142 Estimation unit 143 Learning unit 150 Storage unit 160 Output unit

Claims

1. An information processing method by an information processing apparatus that inputs and outputs information regarding a flow regulating body that regulates the flow within a predetermined space for a predetermined space, comprising: a storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; a reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined; an estimation step of inputting the one-dimensional information received in the reception step into the learning model and estimating the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulating body are set; an output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step; wherein the predetermined space is a space within a container of a deposition reactor that deposits a compound based on a raw material on the surface of a core material, and the container has ejection holes for ejecting the raw material; the flow regulating body is an ejection hole for ejecting the raw material; and the reception step receives the number of the ejection holes included in the plurality of small spaces. An information processing method.

2. The information processing method according to claim 1, wherein the information regarding the flow within the predetermined space is information regarding the heat flux generated around the core material.

3. The information processing method according to claim 1, wherein the information regarding the flow within the predetermined space is information regarding the thickness of the compound deposited on the core material.

4. An information processing method by an information processing apparatus that inputs and outputs information regarding a flow regulating body that regulates the flow within a predetermined space for a predetermined space, comprising: a storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; a reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined; An estimation step of inputting the one-dimensional information received in the reception step into the learning model to estimate the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulator are in place; An output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step; comprising; The predetermined space is the space inside the container of a deposition reactor that deposits a compound based on a raw material on the surface of a core material. The container is provided with an ejection hole for ejecting the raw material and a discharge hole for discharging the gas within the predetermined space. The flow regulator is the discharge hole. The reception step is an information processing method for receiving the number of the discharge holes included in the plurality of small spaces. [

5. ] An information processing method by an information processing apparatus for inputting and outputting information regarding a flow regulator that regulates the flow within a predetermined space, the method comprising: A storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulator included in each small space when defining a plurality of small spaces included in the predetermined space, and information regarding the flow within the predetermined space in that case; A reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulator included in each small space when defining a plurality of small spaces included in the predetermined space; An estimation step of inputting the one-dimensional information received in the reception step into the learning model to estimate the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulator are in place; An output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step; comprising; The predetermined space is the space inside a container filled with a granular material. By supplying gas into the container, the granular material will flow. The flow regulator is an obstruction plate provided at a position within the predetermined space that obstructs the flow in a predetermined direction. The bubbles contained within the predetermined space are split upon contact with the obstruction plate and become smaller in diameter. The information regarding the flow within the predetermined space is information regarding at least either the diameter or the number of the bubbles included in a measurement region within the predetermined space. [

6. ] An information processing method by an information processing apparatus that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space, comprising: A storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; A reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined; An estimation step of inputting the one-dimensional information received in the reception step into the learning model and estimating the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulation body are set; An output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step; Including: The predetermined space is a space inside a container filled with a granular material, has a measurement point for measuring the temperature inside the space, a temperature control means is arranged in the container, and by supplying a gas into the container, the granular material becomes fluid; The flow regulation body is the temperature control means for adjusting the temperature; The information regarding the flow within the predetermined space is information on the temperature of the measurement point within the predetermined space. An information processing method.

7. An information processing method by an information processing apparatus that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space, comprising: A storage step of storing a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; A reception step of receiving an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined; An estimation step of inputting the one-dimensional information received in the reception step into the learning model and estimating the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulation body are set; An output step of outputting information regarding the flow within the predetermined space based on the estimation by the estimation step; comprising; The predetermined space has a measurement area for measuring at least one of temperature and substance concentration within the space, and is a room where gas flows; The flow regulating body is a ventilation opening or an opening provided in the predetermined space; The information regarding the flow within the predetermined space is information regarding at least one of temperature and substance concentration of the measurement area within the predetermined space. An information processing method.

8. An information processing apparatus for inputting and outputting information regarding a flow regulating body that regulates the flow within a predetermined space for the predetermined space, A storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; A reception unit that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined; An estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulating body are indicated by the one-dimensional information; An output unit that outputs information regarding the flow within the predetermined space based on the estimation by the estimation unit; comprising; The predetermined space is the space inside the container of a deposition reactor that deposits a compound based on a raw material on the surface of a core material, and the container has ejection holes for ejecting the raw material; The flow regulating body is the ejection holes for ejecting the raw material; The reception unit is an information processing apparatus that receives the number of the ejection holes included in the plurality of small spaces.

9. A computer for inputting and outputting information regarding a flow regulating body that regulates the flow within a predetermined space for the predetermined space, to a computer that can access a storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case; A reception function that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the fluid regulation bodies included in each small space when defining a plurality of small spaces included in the predetermined space; An estimation function that inputs the one-dimensional information received by the reception function into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the fluid regulation bodies indicated by the one-dimensional information are set; An output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function; To realize; The predetermined space is the space inside a container of a precipitation reactor that precipitates a compound based on a raw material on the surface of a core material, and the container has ejection holes for ejecting the raw material. The fluid regulation body is an ejection hole for ejecting the raw material. The reception function is an information processing program that receives the number of the ejection holes included in the plurality of small spaces.

10. An information processing apparatus that inputs and outputs information regarding a fluid regulation body that regulates the flow within a predetermined space for the predetermined space, A storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the fluid regulation bodies included in each small space when defining a plurality of small spaces included in the predetermined space, and information regarding the flow within the predetermined space in that case; A reception unit that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the fluid regulation bodies included in each small space when defining a plurality of small spaces included in the predetermined space; An estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the fluid regulation bodies indicated by the one-dimensional information are set; An output unit that outputs information regarding the flow within the predetermined space based on the estimation by the estimation unit; Comprising: The predetermined space is the space inside a container of a precipitation reactor that precipitates a compound based on a raw material on the surface of a core material. The container includes ejection holes for ejecting the raw material and discharge holes for discharging gas within the predetermined space. The fluid regulation body is the discharge hole. The reception unit receives the number of the discharge holes included in the plurality of small spaces. Information processing apparatus.

11. A computer that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space, the computer being accessible to a storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case. A reception function that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined. An estimation function that inputs the one-dimensional information received by the reception function into the learning model and estimates the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulation body are set. An output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function. To implement The predetermined space is the space inside a container of a deposition reactor that deposits a compound based on a raw material on the surface of a core material. The container includes an ejection hole for ejecting the raw material and a discharge hole for discharging the gas within the predetermined space. The flow regulation body is the discharge hole. The reception function receives the number of the discharge holes included in the plurality of small spaces. Information processing program.

12. An information processing method by an information processing apparatus that inputs and outputs information regarding a flow regulation body that regulates the flow within a predetermined space, the method including: A storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined, and information regarding the flow within the predetermined space in that case. A reception unit that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulation body included in each small space when a plurality of small spaces included in the predetermined space are defined. An estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulation body are set. An output unit that outputs information regarding the flow within the predetermined space based on the estimation by the estimation unit. Comprising The predetermined space is the space inside a container filled with powder or granular material, and by supplying gas into the container, the powder or granular material becomes fluid, the flow regulating body is an obstruction plate provided at a position in the predetermined space that obstructs the flow in a predetermined direction, the air bubbles contained in the space within the predetermined space split upon contact with the obstruction plate and their diameters become smaller, information regarding the flow within the predetermined space is information regarding at least either the diameter or the number of air bubbles contained in a measurement region within the predetermined space, and is an information processing apparatus.

13. A computer that inputs and outputs information regarding a flow regulating body that regulates the flow within a predetermined space, and that can access a storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space and information regarding the flow regulating body contained in each small space when a plurality of small spaces contained in the predetermined space are defined, and information regarding the flow within the predetermined space in that case. The computer has a reception function that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating body contained in each small space when a plurality of small spaces contained in the predetermined space are defined, inputs the one-dimensional information received by the reception function into the learning model, and has an estimation function that estimates the flow within the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulating body are as described, and has an output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function, and is realized, the predetermined space is the space inside a container filled with powder or granular material, by supplying gas into the container, the powder or granular material becomes fluid, the flow regulating body is an obstruction plate provided at a position in the predetermined space that obstructs the flow in a predetermined direction, the air bubbles contained in the space within the predetermined space split upon contact with the obstruction plate and their diameters become smaller, information regarding the flow within the predetermined space is information regarding at least either the diameter or the number of air bubbles contained in a measurement region within the predetermined space, and is an information processing program.

14. An information processing apparatus that inputs and outputs information regarding a flow regulating body that regulates the flow within a predetermined space, A storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating bodies included in each small space when defining a plurality of small spaces included in the predetermined space, and information regarding the flow within the predetermined space in that case; A reception unit that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating bodies included in each small space when defining a plurality of small spaces included in the predetermined space; An estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulating bodies indicated by the one-dimensional information are set; An output unit that outputs information regarding the flow within the predetermined space based on the estimation by the estimation unit; comprising; The predetermined space is the space inside a container filled with powder particles, has a measurement point for measuring the temperature inside the space, a temperature control means is arranged in the container, and by supplying gas into the container, the powder particles are caused to flow; The flow regulating body is the temperature control means for adjusting the temperature; The information regarding the flow within the predetermined space is information on the temperature of the measurement point within the predetermined space. An information processing device.

15. A computer that inputs and outputs information regarding a flow regulating body that regulates the flow within a predetermined space, the computer being accessible to a storage unit that stores a learning model that has learned the relationship between information representing the state of the predetermined space, information regarding the flow regulating bodies included in each small space when defining a plurality of small spaces included in the predetermined space, and information regarding the flow within the predetermined space in that case; A reception function that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating bodies included in each small space when defining a plurality of small spaces included in the predetermined space; An estimation function that inputs the one-dimensional information received by the reception function into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulating bodies indicated by the one-dimensional information are set; An output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function; is realized; The predetermined space is the space inside a container filled with powder or granular material, has a measurement point for measuring the temperature inside the space, a temperature control means is arranged in the container, and by supplying gas into the container, the powder or granular material becomes fluid, The flow regulating body is the temperature control means for temperature adjustment, The information regarding the flow in the predetermined space is an information processing program which is the temperature of the measurement point in the predetermined space.

16. An information processing apparatus for inputting and outputting information regarding a flow regulating body that regulates the flow in a predetermined space, A storage unit that stores a learning model which has learned the relationship between the information representing the state of the predetermined space, the information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and the information regarding the flow in the predetermined space in that case, A reception unit that receives an input of one-dimensional information indicating the information representing the state of the predetermined space and the information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, An estimation unit that inputs the one-dimensional information received by the reception unit into the learning model and estimates the flow in the predetermined space when the state of the predetermined space indicated by the one-dimensional information and the arrangement of the flow regulating body are set, An output unit that outputs the information regarding the flow in the predetermined space based on the estimation by the estimation unit, comprising, The predetermined space has a measurement area for measuring at least one of the temperature and the concentration of the substance inside the space, and is a room where gas flows, The flow regulating body is a ventilation opening or an opening provided in the predetermined space, The information processing apparatus, wherein the information regarding the flow in the predetermined space is information regarding at least one of the temperature and the concentration of the substance in the measurement area in the predetermined space.

17. A computer for inputting and outputting information regarding a flow regulating body that regulates the flow in a predetermined space, to a computer that can access a storage unit that stores a learning model which has learned the relationship between the information representing the state of the predetermined space, the information regarding the flow regulating body included in each small space when a plurality of small spaces included in the predetermined space are defined, and the information regarding the flow in the predetermined space in that case, A reception function that receives an input of one-dimensional information indicating information representing the state of the predetermined space and information regarding the flow regulating bodies included in each small space when a plurality of small spaces included in the predetermined space are defined; An estimation function that inputs the one-dimensional information received by the reception function into the learning model and estimates the flow within the predetermined space when the state of the predetermined space and the arrangement of the flow regulating bodies indicated by the one-dimensional information are set; An output function that outputs information regarding the flow within the predetermined space based on the estimation by the estimation function; is realized, The predetermined space has a measurement area that measures at least one of temperature and substance concentration within the space, and is a room in which gas flows, The flow regulating body is a ventilation opening or an opening provided in the predetermined space, An information processing program in which the information regarding the flow within the predetermined space is information regarding at least one of temperature and substance concentration of the measurement area within the predetermined space.

Images