Assistance device, assistance method, and assistance program

Abstract

The present invention provides an assistance device, an assistance method, and an assistance program that are for assisting operation of a crystal growth furnace. An assistance device according to the present invention comprises: a calculation unit that outputs a temperature distribution of a crystal surface on the basis of temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device which heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and a display unit that displays the temperature distribution of the crystal surface output by the calculation unit.

Description

Support device, support method, and support program

[0001] This disclosure relates to a support device, a support method, and a support program.

[0002] As a technique for analyzing the crystal growth process in a crystal growth furnace, simulation techniques that combine temperature field calculations with fluid and chemical reaction calculations are known. This simulation technique reproduces the crystal growth process by repeatedly predicting the state of the crystal after time Δt, and therefore, it is possible to analyze in detail how the crystal growth process changes when manufacturing conditions are changed.

[0003] Patent No. 7472969

[0004] Yoshiki Inoue et al., "Optimization of Furnace Temperature Distribution using Machine Learning in SiC Sublimation Method," The 82nd Autumn Meeting of the Japan Society of Applied Physics, September 10-13, 2021.

[0005] However, the above simulation technology has a problem in that it requires a large amount of computation, resulting in an enormous execution time (for example, it takes about the same amount of time as it takes to grow crystals in an actual crystal growth furnace). For this reason, with the above simulation technology, it is difficult to realize operational support, such as running the process while changing the manufacturing conditions in an operating crystal growth furnace and reflecting the appropriate manufacturing conditions found based on the execution results in the operating crystal growth furnace in real time.

[0006] This disclosure provides support devices, support methods, and support programs for assisting the operation of a crystal growth furnace.

[0007] A support device according to a first aspect of the present disclosure includes a calculation unit that outputs a temperature distribution of the crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace, and a display unit that displays the temperature distribution of the crystal surface output by the calculation unit.

[0008] A second aspect of this disclosure is the support device described in the first aspect, wherein if information indicating the position or output of the heating device is changed in response to the display of the temperature distribution of the crystal surface, the calculation unit recalculates based on the changed information, and the display unit updates the display of the temperature distribution of the crystal surface based on the recalculation result.

[0009] A third aspect of the present disclosure is a support device according to the first or second aspect, wherein the calculation unit has a physical model in which information indicating the structure of the crystal growth furnace, physical properties of each component, boundary conditions between each component, and information indicating the chemical structure of the raw material are set, and the physical model calculates the temperature field inside the crystal growth furnace by inputting temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and the position of the crystal surface in the crystal growth furnace.

[0010] A fourth aspect of the present disclosure is a support device according to any one of the first to third aspects, wherein the calculation unit has a surrogate model trained using training data in which temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace are explanatory variables, and the temperature distribution of the crystal surface is the objective variable.

[0011] A fifth aspect of the present disclosure is a support device according to any of the first to fourth aspects, wherein the display unit displays a display screen including: an input area for inputting temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device for heating the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and an output area for displaying the temperature distribution of the crystal surface.

[0012] A support method according to a sixth aspect of this disclosure involves a computer performing a calculation step of calculating the temperature distribution of the crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and a display step of displaying the temperature distribution of the crystal surface calculated in the calculation step.

[0013] A support program according to a seventh aspect of this disclosure causes a computer to perform a calculation step of calculating the temperature distribution of the crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace, and a display step of displaying the temperature distribution of the crystal surface calculated in the calculation step.

[0014] According to this disclosure, it is possible to provide support devices, support methods, and support programs for supporting the operation of a crystal growth furnace.

[0015] Figure 1 is a diagram showing an example of application of the support device according to the first embodiment. Figure 2 is a diagram showing an example of the support mode of the support device when applied. Figure 3 is a diagram showing an example of the hardware configuration of the support device. Figure 4 is a diagram for explaining general simulation technology. Figure 5 is a diagram for explaining the simulation technology installed in the support device. Figure 6 is a diagram showing an example of the functional configuration of the support device. Figure 7 is a diagram showing an example of the display screen of the support device. Figure 8 is a first flowchart showing the flow of support processing by the support device. Figure 9 is a diagram showing an example of application of the support device according to the second embodiment. Figure 10 is a diagram showing an example of training data. Figure 11 is a diagram showing an example of the functional configuration of the learning device. Figure 12 is a flowchart showing the flow of learning processing by the temperature field calculation device and the learning device. Figure 13 is a diagram showing an example of the functional configuration of the support device. Figure 14 is a second flowchart showing the flow of support processing by the support device.

[0016] Each embodiment will be described below with reference to the attached drawings. In this specification and the drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant descriptions will be omitted.

[0017] [First Embodiment] <Example of Application of the Support Device> First, an example of application of the support device according to the first embodiment will be described. Figure 1 is a diagram showing an example of application of the support device according to the first embodiment.

[0018] As shown in Figure 1, the support device 140 according to the first embodiment is applied to, for example, a crystal manufacturing system 10 and is operated by an operator 160 who operates the crystal manufacturing system 10 during the operation of the crystal growth furnace 110.

[0019] The crystal manufacturing system 10 comprises a crystal growth furnace 110, a control device 120, and an operating device 130. The crystal growth furnace 110 is a device that manufactures crystals by feeding in raw materials and seed crystals under predetermined manufacturing conditions. The crystal growth furnace 110 is a furnace in which crystal growth cannot be visually observed from the outside during operation, and is a heating furnace that manufactures crystals in which the temperature distribution affects growth. The crystals referred to here include both single crystals and polycrystalline crystals, and specific examples of single crystals include Si single crystals and SiC single crystals.

[0020] The control device 120 controls the crystal growth furnace 110 so that it operates under set manufacturing conditions. The manufacturing conditions set in the control device 120 include, for example, the temperature inside the crystal growth furnace 110, the position and output of the heating device that heats the crystal growth furnace 110, and so on.

[0021] The operating device 130 receives input of predetermined manufacturing conditions from the operator 160, notifies the control device 120, and displays the temperature data inside the crystal growth furnace 110, acquired by the control device 120, to the operator 160.

[0022] The support device 140 performs calculations for the crystal growth furnace 110, for example, based on instructions from the operator 160. Specifically, using the user interface device 150, the support device 140 calculates the temperature field inside the crystal growth furnace 110 when the operator 160 inputs the information necessary to calculate the temperature field inside the crystal growth furnace 110.

[0023] The support device 140 derives the temperature distribution of the crystal surface inside the crystal growth furnace 110 based on the temperature field calculation results and displays it on the user interface device 150. The operator 160 changes the manufacturing conditions based on the temperature distribution of the crystal surface displayed on the user interface device 150 and operates the control device 130 so that a crystal of the appropriate shape (for example, a crystal with no radial irregularities and a flat surface) is manufactured.

[0024] As described above, the support device 140 according to the first embodiment derives the temperature distribution of the crystal surface in real time during the operation of the crystal growth furnace 110 and displays it to the operator 160. This allows the operator 160 to search for appropriate manufacturing conditions and reflect them in real time in the crystal growth furnace during operation. In other words, the support device 140 according to the first embodiment can support the operation of the crystal growth furnace 110.

[0025] <Specific Examples of Support Modes> Next, specific examples of support modes used by the support device 140 applied to the crystal manufacturing system 10 to support the operation of the crystal growth furnace 110 will be described. Figure 2 shows an example of the support mode of the support device when applied.

[0026] As shown in Figure 2, once preconditions are set for the support device 140 and manufacturing conditions are set for the control device 120, the operator 160 starts operating the crystal growth furnace 110. This starts the production of crystals by the crystal growth furnace 110, and the control device 120 begins monitoring and controlling the crystal growth furnace 110. The preconditions set for the support device 140 are the information set in the physical model (a physical model that calculates the temperature field inside the crystal growth furnace 110) of the support device 140, and include, for example: information indicating the structure of the crystal growth furnace 110, material properties of each component constituting the crystal growth furnace 110, boundary conditions between each component constituting the crystal growth furnace 110, and information indicating the chemical structure of the raw materials fed into the crystal growth furnace 110.

[0027] During operation of the crystal growth furnace 110, the operator 160 uses the support device 140 to calculate the temperature field inside the crystal growth furnace 110. Specifically, the operator 160 inputs the following to the support device 140: - Temperature data inside the crystal growth furnace 110 measured during operation; - Information indicating the current position of the heating device in the crystal growth furnace 110 during operation; - Information indicating the current output of the heating device in the crystal growth furnace 110 during operation; and - Information indicating the current position (or amount of crystal growth) of the crystal surface inside the crystal growth furnace 110. Based on this, the support device 140 calculates the temperature field inside the crystal growth furnace 110 and displays the derived temperature distribution of the crystal surface inside the crystal growth furnace 110 to the operator 160. Here, the position of the crystal surface may be, for example, the center position of the crystal surface, or the average position of multiple positions on the crystal surface.

[0028] If the displayed temperature distribution on the crystal surface is uniform, the operator 160 continues operation of the crystal growth furnace 110 without changing the manufacturing conditions.

[0029] On the other hand, if there is variation in the displayed temperature distribution of the crystal surface, operator 160 considers changing the position or output of the heating device to suppress the variation in the temperature distribution of the crystal surface. Operator 160 inputs the changed position or output of the heating device to the support device 140. As a result, the support device 140 recalculates the temperature field inside the crystal growth furnace 110. Based on the recalculation results, the support device 140 updates the temperature distribution of the crystal surface inside the crystal growth furnace 110 and displays it to operator 160.

[0030] If the updated temperature distribution on the crystal surface based on the recalculation results is not uniform, further changes to the position or output of the heating device are considered. Operator 160 inputs the further changed position or output of the heating device to the support device 140. Operator 160 repeats the operation until the updated temperature distribution on the crystal surface based on the recalculation results becomes uniform.

[0031] If the updated temperature distribution on the crystal surface, based on the recalculation results, becomes uniform, operator 160 sets the manufacturing conditions, including the changed position or output of the heating device, to the control device 120.

[0032] Thus, with the above support configuration, the operator 160 can check the temperature distribution of the crystal surface in real time by running a physical model during the operation of the crystal growth furnace 110. This allows the operator 160 to search for appropriate manufacturing conditions and reflect the searched appropriate manufacturing conditions in the crystal growth furnace 110 in real time.

[0033] <Hardware Configuration of the Support Device> Next, the hardware configuration of the support device 140 will be described. Figure 3 shows an example of the hardware configuration of the support device.

[0034] As shown in Figure 3, the support device 140 includes a processor 301, memory 302, auxiliary storage device 303, I / F (Interface) device 304, communication device 305, and drive device 306. The hardware of the support device 140 is interconnected via bus 307.

[0035] The processor 301 has various arithmetic devices such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit). The processor 301 executes by reading various programs (for example, support programs, etc.) onto the memory 302.

[0036] The memory 302 has main memory devices such as a ROM (Read Only Memory) and a RAM (Random Access Memory). The processor 301 and the memory 302 form a so-called computer, and the computer realizes various functions by the processor 301 executing various programs read onto the memory 302.

[0037] The auxiliary storage device 303 stores various programs and various data used when the various programs are executed by the processor 301.

[0038] The I / F device 304 is a connection device for connecting the user interface device 150. The communication device 305 is a communication device for communicating with an external device via a network (not shown).

[0039] The drive device 306 is a device for setting the recording medium 312. The recording medium 312 here includes media that record information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, a magneto-optical disk, etc. The recording medium 312 may include semiconductor memories that record information electrically, such as a ROM, a flash memory, etc.

[0040] Note that the various programs installed in the auxiliary storage device 303 are installed, for example, when the distributed recording medium 312 is set in the drive device 306 and the various programs recorded on the recording medium 312 are read by the drive device 306. Alternatively, the various programs installed in the auxiliary storage device 303 may be installed when downloaded from a network via the communication device 305.

[0041] <Explanation of Simulation Technology> Next, the simulation technology installed in the support device 140 will be explained. In this explanation, first, the general simulation technology for analyzing the crystal growth process in the crystal growth furnace will be explained to clarify the characteristics of the simulation technology installed in the support device 140.

[0042] (1) General Simulation Technology FIG. 4 is a diagram for explaining the general simulation technology. As shown in FIG. 4, generally, when analyzing the crystal growth process in the crystal growth furnace, a temperature field calculation device 410 and a fluid and chemical reaction calculation device 420 are used.

[0043] The temperature field calculation device 410 is a device that calculates the temperature field in the crystal growth furnace. When the temperature field calculation device 410 calculates the temperature field in the crystal growth furnace, as prerequisite conditions, - information indicating the structure of the crystal growth furnace, - material physical property values of each member constituting the crystal growth furnace, - boundary conditions between each member constituting the crystal growth furnace, - information indicating the chemical structure of the raw material input into the crystal growth furnace, are set, and as manufacturing conditions, - temperature data measured in the crystal growth furnace, - information indicating the position of the heating device for heating the crystal growth furnace, - information indicating the output of the heating device for heating the crystal growth furnace, are input. Thereby, the temperature field calculation device 410 outputs the temperature distribution in the crystal growth furnace.

[0044] The fluid and chemical reaction calculation device 420 is a device that calculates the fluid and chemical reactions. When the fluid and chemical reaction calculation device 420 calculates the fluid and chemical reactions, as prerequisite conditions, - information indicating the structure of the crystal growth furnace, - material physical property values of each member constituting the crystal growth furnace, - boundary conditions between each member constituting the crystal growth furnace, - information indicating the chemical structure of the raw material input into the crystal growth furnace, are set, and as input data, - the temperature distribution in the crystal growth furnace output by the temperature field calculation device 410, is input. Thereby, the fluid and chemical reaction calculation device 420 outputs the crystal growth amount (or the position of the crystal surface in the crystal growth furnace) as the state of the crystal after Δt time.

[0045] The crystal growth amount output by the fluid and chemical reaction calculation device 420 is fed back to the temperature field calculation device 410. As a result, the preconditions are updated, and the updated preconditions are set for the temperature field calculation device 410, which then outputs the temperature distribution in the crystal growth furnace after Δt time.

[0046] Thus, in general simulation techniques, the crystal growth process is reproduced by repeatedly combining the following: • Calculation of the temperature field by the temperature field calculation device 410, and • Calculation of fluid and chemical reactions by the fluid and chemical reaction calculation device 420, and outputting the state of the crystal after Δt time.

[0047] Therefore, the implementation would take an enormous amount of time (for example, about the same amount of time required to grow crystals in an actual crystal growth furnace), making it difficult to realize the operational support shown in Figure 2.

[0048] (2) Simulation technology installed in the support device 140 Figure 5 is a diagram for explaining the simulation technology installed in the support device. The differences from general simulation technology are: - In the case of general simulation technology, the amount of crystal growth is output, whereas in the case of the simulation technology installed in the support device 140, the temperature distribution of the crystal surface inside the crystal growth furnace is output. - In other words, in the case of the simulation technology installed in the support device 140, fluid and chemical reaction calculations are not performed. - In the case of general simulation technology, the amount of crystal growth after Δt time is fed back and the preconditions are updated, whereas in the case of the simulation technology installed in the support device 140, no feedback processing is performed and the operator 160 inputs the amount of crystal growth. - In other words, in the case of the simulation technology installed in the support device 140, iterative processing is not performed. As a result, the execution time can be significantly reduced in the case of the simulation technology installed in the support device 140.

[0049] Another difference from general simulation techniques is that, while general simulation techniques display changes in crystal growth when manufacturing conditions are changed, the support device 140 displays the temperature distribution on the crystal surface when manufacturing conditions are changed. As a result, the simulation technique installed in the support device 140 can display information that is useful for searching for appropriate manufacturing conditions, rather than information that is useful for analyzing the crystal growth process.

[0050] As a result, the support device 140 can provide the operational support shown in Figure 2.

[0051] <Functional Configuration of the Support Device> Next, the functional configuration of the support device 140 will be described. Figure 6 is a diagram showing an example of the functional configuration of the support device. The support device 140 has a support program installed, and when this program is executed, the support device 140 functions as a prerequisite setting unit 610, a manufacturing condition acquisition unit 620, a calculation unit 630, and a display unit 640, as shown in Figure 6.

[0052] The precondition setting unit 610 receives preconditions to be set in the physical model (in the example in Figure 6, the temperature field calculation unit 631 of the calculation unit 630) for calculating the temperature field inside the crystal growth furnace 110 from the setter (not shown) of the support device 140. The precondition setting unit 610 sets the received preconditions in the temperature field calculation unit 631 of the calculation unit 630.

[0053] The manufacturing condition acquisition unit 620 acquires the manufacturing conditions necessary to operate the physical model (in the example in Figure 6, the temperature field calculation unit 631 of the calculation unit 630) that calculates the temperature field inside the crystal growth furnace 110. The manufacturing condition acquisition unit 620 acquires the manufacturing conditions entered into the display screen provided by the display unit 640. The manufacturing condition acquisition unit 620 acquires the following as manufacturing conditions: - Temperature data inside the crystal growth furnace 110 measured during the operation of the crystal growth furnace 110; - Information indicating the current position of the heating device of the crystal growth furnace 110 during operation; - Information indicating the current output of the heating device of the crystal growth furnace 110 during operation; and - Information indicating the current position of the crystal surface (or crystal growth amount) inside the crystal growth furnace 110. The manufacturing condition acquisition unit 620 inputs the acquired manufacturing conditions into the temperature field calculation unit 631 of the calculation unit 630.

[0054] The calculation unit 630 includes a temperature field calculation unit 631 and a surface temperature derivation unit 632. The temperature field calculation unit 631 is a physical model, and when manufacturing conditions are input by the manufacturing condition acquisition unit 620, the temperature field calculation unit 631 is operated under the set preconditions to calculate the temperature field inside the crystal growth furnace 110. The surface temperature derivation unit 632 derives the temperature distribution of the crystal surface inside the crystal growth furnace 110 based on the temperature field calculation result (temperature distribution inside the crystal growth furnace 110) by the temperature field calculation unit 631.

[0055] The display unit 640 displays a display screen on the user interface device 150 that includes an input area for receiving input of manufacturing conditions and an output area for displaying the temperature distribution of the crystal surface derived by the surface temperature derivation unit 632.

[0056] <Display screen of the support device> Next, a specific example of the display screen displayed on the user interface device 150 of the support device 140 during the operation of the crystal growth furnace 110 will be described. Figure 7 shows an example of the display screen of the support device.

[0057] As shown in Figure 7, the display screens 710 and 720 include an input area for receiving input of manufacturing conditions and an output area for displaying the temperature distribution of the crystal surface. As described above, the manufacturing conditions include: - Temperature data inside the crystal growth furnace 110 measured during the operation of the crystal growth furnace 110; - Information indicating the current position of the heating device of the crystal growth furnace 110 during operation; - Information indicating the current output of the heating device of the crystal growth furnace 110 during operation; and - Information indicating the current position (or amount of crystal growth) of the crystal surface inside the crystal growth furnace 110. For this reason, the input areas of the display screens 710 and 720 are provided with sliders for inputting values ​​for the corresponding items. In the output areas of the display screens 710 and 720, the horizontal axis represents each position on the crystal surface, and the vertical axis represents temperature. In other words, the graph shown in the output areas of the display screens 710 and 720 is a one-dimensional representation of the temperature distribution of the crystal surface. A one-dimensional temperature distribution on a crystal surface refers to the temperature distribution in the radial direction of the circle, for example, when the crystal surface can be considered to be approximately a circle.

[0058] Operator 160 checks the temperature distribution of the crystal surface displayed in the output area, and if the temperature distribution is not uniform, moves a slider to change the position or output of the heating device, thereby causing the temperature field inside the crystal growth furnace 110 to be recalculated. As a result, the temperature distribution of the crystal surface displayed in the output area of ​​the display screen 720 is updated based on the recalculation results. In this way, by operating the display screen 710, operator 160 can search for the appropriate position or output of the heating device to make the temperature distribution of the crystal surface uniform.

[0059] In Figure 7, the display screen 710 shows the uneven temperature distribution on the crystal surface before the appropriate heating device position or output is found. On the other hand, in Figure 7, the display screen 720 shows the uniform temperature distribution on the crystal surface, which has been updated based on the recalculation results after the appropriate heating device position or output has been found.

[0060] <Flow of support processing by the support device> Next, the flow of support processing by the support device 140 will be explained. Figure 9 is the first flowchart showing the flow of support processing by the support device.

[0061] In step S801, the support device 140 receives the setting of preconditions by the person who sets the support device 140, and sets the received preconditions in the temperature field calculation unit 631.

[0062] In step S802, the support device 140 displays a screen and accepts input of manufacturing conditions from the operator 160. The support device 140 inputs the received manufacturing conditions into the temperature field calculation unit 631.

[0063] In step S803, the support device 140 operates the calculation unit 630. Specifically, the temperature field calculation unit 631 calculates the temperature field inside the crystal growth furnace, and the surface temperature derivation unit 632 derives the temperature distribution of the crystal surface.

[0064] In step S804, the support device 140 displays the temperature distribution of the crystal surface on the display screen.

[0065] In step S805, the support device 140 determines on the display screen whether the position of the heating device or the output of the heating device in the crystal growth furnace 110 has been changed. If it is determined in step S805 that the position of the heating device or the output of the heating device has been changed (if the answer is YES in step S805), the process returns to step S803. On the other hand, if it is determined in step S805 that the position of the heating device and the output of the heating device have not been changed (if the answer is NO in step S805), the process proceeds to step S806.

[0066] In step S806, the support device 140 determines whether or not to continue the support process. If it determines in step S806 to continue the support process (if the answer is YES in step S806), it returns to step S802. On the other hand, if it determines not to continue the support process (if the answer is NO in step S806), it terminates the support process.

[0067] <Summary> As is clear from the above explanation, the support device 140 according to the first embodiment calculates the temperature field inside the crystal growth furnace 110 and derives the temperature distribution of the crystal surface based on the temperature data measured in the crystal growth furnace 110, information indicating the position and output of the heating device that heats the crystal growth furnace 110, and information indicating the position of the crystal surface in the crystal growth furnace 110. It also displays the derived temperature distribution of the crystal surface.

[0068] As described above, the support device 140 according to the first embodiment operates a physical model during the operation of the crystal growth furnace and displays the temperature distribution of the crystal surface to the operator in real time. As a result, the support device 140 according to the first embodiment enables the operator to search for appropriate manufacturing conditions and to reflect the searched appropriate manufacturing conditions in the crystal growth furnace, thereby supporting the operation of the crystal growth furnace.

[0069] [Second Embodiment] The support device 140 according to the first embodiment described above is configured to operate a physical model in order to derive the temperature distribution of the crystal surface. However, the model operated in order to derive the temperature distribution of the crystal surface is not limited to a physical model. For example, a surrogate model may be operated which is constructed by generating training data by operating the physical model and then training using the training data. The second embodiment will be described below, focusing on the differences from the first embodiment described above.

[0070] <Examples of Application of the Support Device> First, we will describe examples of application of the support device according to the second embodiment. Figure 9 is a diagram showing an example of application of the support device according to the second embodiment. The differences from the example of application described using Figure 1 in the first embodiment are that the function of the support device 910 is different from the function of the support device 140 in Figure 1, and that it has a temperature field calculation device 410 and a learning device 920.

[0071] The temperature field calculation device 410 has the same functions as the temperature field calculation device 410 described with reference to Figure 4 in the first embodiment, for example. As described in the first embodiment, when the temperature field calculation device 410 calculates the temperature field inside the crystal growth furnace, the following preconditions are set: information indicating the structure of the crystal growth furnace, material properties of each component constituting the crystal growth furnace, boundary conditions between each component constituting the crystal growth furnace, and information indicating the chemical structure of the raw material fed into the crystal growth furnace. In the case of the second embodiment, the temperature field calculation device 410 is input as manufacturing conditions: temperature data measured in the crystal growth furnace, information indicating the position of the heating device that heats the crystal growth furnace, information indicating the output of the heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface (or amount of crystal growth) inside the crystal growth furnace. Based on this, the temperature field calculation device 410 calculates the temperature field inside the crystal growth furnace.

[0072] In the second embodiment, various manufacturing conditions for learning are input to the temperature field calculation device 410, and the temperature field inside the crystal growth furnace is calculated. The temperature field calculation device 410 generates learning data by associating the various input manufacturing conditions with the temperature distributions of various crystal surfaces derived from the calculated temperature field inside the crystal growth furnace. The learning data generated by the temperature field calculation device 410 is notified to the learning device 920.

[0073] The learning device 920 uses the training data notified by the temperature field calculation device 410 to train a machine learning model (for example, a Neural Network (NN)) and construct a surrogate model.

[0074] The support device 910 has a surrogate model constructed by the learning device 920. When information necessary for calculating the temperature distribution on the crystal surface is input by the operator 160 using, for example, the user interface device 150, the support device 910 operates the surrogate model to calculate the temperature distribution on the crystal surface. The support device 910 displays the calculated temperature distribution on the crystal surface on the user interface device 150. Based on the temperature distribution on the crystal surface displayed on the user interface device 150, the operator 160 changes the manufacturing conditions and operates the operating device 130 so that a crystal with an appropriate shape (for example, a crystal without irregularities in the radial direction and with a flat surface) is manufactured.

[0075] <An Example of Learning Data> Next, a specific example of learning data will be described. FIG. 10 is a diagram showing an example of learning data. As shown in FIG. 10, the learning data 1000 includes "manufacturing conditions" and "temperature distribution on the crystal surface" as items of information.

[0076] Various manufacturing conditions input to the temperature field calculation device 410 are stored in the "manufacturing conditions". The example in FIG. 10 shows that as temperature data measured in the crystal growth furnace 110, "sensor temperature T １ ", as information indicating the position of the heating device that heats the crystal growth furnace 110, "heating device position Y １ ", as information indicating the output of the heating device that heats the crystal growth furnace 110, "heating device output V １ ", and as information indicating the position of the crystal surface in the crystal growth furnace 110, "position of crystal surface Z １ " are stored.

[0077] The "temperature distribution on the crystal surface" stores the temperature distribution on the crystal surface derived by calculating the temperature field in the crystal growth furnace 110 when the manufacturing conditions are input to the temperature field calculation device 410 with pre-set conditions. The example in FIG. 10 shows that the temperatures at each coordinate (x １ , y １ ) to (x ｎ , y ｎ ) on the crystal surface are t １１ to t ｎ１ .

[0078] In this way, training data 1000 is generated by associating the "manufacturing conditions" (explanatory variables) and the "temperature distribution on the crystal surface" (target variable) when the temperature field calculation device 410 is operated under various manufacturing conditions.

[0079] <Functional Configuration of the Learning Device> Next, the functional configuration of the learning device 920 will be described. Figure 11 is a diagram showing an example of the functional configuration of the learning device. The learning device 920 has a learning program installed, and when this program is executed, the learning device 920 functions as a learning unit 1100.

[0080] The learning unit 1100 includes a machine learning model 1110 and a comparison / modification unit 1120. The machine learning model 1110 calculates the temperature distribution of the crystal surface when the manufacturing conditions (explanatory variables) of the training data 1000 are input. The comparison / modification unit 1120 compares the temperature distribution of the crystal surface calculated by the machine learning model 1110 with the temperature distribution of the crystal surface stored in the training data 1000 (target variable), and updates the model parameters of the machine learning model 1110 based on the error between the two. As a result, the learning unit 1100 constructs a surrogate model.

[0081] <Learning Process Flow> Next, the learning process flow by the temperature field calculation device 410 and the learning device 920 will be explained. Figure 12 is a flowchart showing the learning process flow by the temperature field calculation device and the learning device.

[0082] In step S1201, the temperature field calculation device 410 accepts the setting of preconditions.

[0083] In step S1202, the temperature field calculation device 410 accepts input of various manufacturing conditions.

[0084] In step S1203, the temperature field calculation device 410 receives various manufacturing conditions as input and calculates the temperature field inside the crystal growth furnace under the set preconditions, and derives the temperature distribution of the crystal surface.

[0085] In step S1204, the temperature field calculation device 410 determines whether the amount of data on the temperature distribution of the crystal surface derived exceeds a predetermined amount of data (a sufficient amount of data for machine learning). If it is determined in step S1204 that the predetermined amount of data has not been exceeded (if the answer is NO in step S1204), the process returns to step S1202. On the other hand, if it is determined in step S1204 that the predetermined amount of data has been exceeded (if the answer is YES in step S1204), the process proceeds to step S1205.

[0086] In step S1205, the temperature field calculation device 410 generates training data by associating the "manufacturing conditions" with the "temperature distribution on the crystal surface".

[0087] In step S1206, the learning device 920 constructs a surrogate model by training a machine learning model using the generated training data.

[0088] In step S1207, the learning device 920 stores the constructed surrogate model in the support device 910.

[0089] <Functional Configuration of the Support Device> Next, the functional configuration of the support device 910 according to the second embodiment will be described. Figure 13 is a diagram showing an example of the functional configuration of the support device. As described above, a support program is installed in the support device 910. In the second embodiment, when the support program is executed, the support device 910 functions as a manufacturing condition acquisition unit 620, a calculation unit 1310, and a display unit 640.

[0090] Of these, the manufacturing condition acquisition unit 620 and the display unit 640 have already been explained using Figure 6 in the first embodiment described above, so their explanation will be omitted here.

[0091] The calculation unit 1310 has a surrogate model 1311. The surrogate model 1311 calculates the temperature distribution on the crystal surface when manufacturing conditions are input.

[0092] <Flow of Support Processing> Next, the flow of support processing by the support device 910 will be explained. Figure 14 is a second flowchart showing the flow of support processing by the support device. The difference from the first flowchart shown in Figure 8 is that it does not include step S801 and includes step S1401 instead of step S803.

[0093] In step S1401, the support device 910 operates the calculation unit 1310. Specifically, the surrogate model 1311 calculates the temperature distribution on the crystal surface based on the input manufacturing conditions.

[0094] <Summary> As is clear from the above explanation, the support device 910 according to the second embodiment: - Calculates the temperature distribution of the crystal surface inside the crystal growth furnace 110 based on the temperature data measured in the crystal growth furnace 110, the position and output of the heating device that heats the crystal growth furnace 110, and the position of the crystal surface inside the crystal growth furnace 110. - Has a display unit 640 that displays the calculated temperature distribution of the crystal surface.

[0095] As described above, the support device 910 according to the second embodiment operates a surrogate model during the operation of the crystal growth furnace and displays the temperature distribution of the crystal surface to the operator in real time. As a result, the support device 910 according to the second embodiment enables the operator to search for appropriate manufacturing conditions and to reflect the searched appropriate manufacturing conditions in the crystal growth furnace, thereby supporting the operation of the crystal growth furnace.

[0096] [Other Embodiments] In the above embodiments, the case in which the position of the crystal surface (or the amount of crystal growth) is input as a manufacturing condition has been described, but the shape of the crystal surface may be input instead of the position of the crystal surface. In other words, in the above embodiments, the position of the crystal surface (or the amount of crystal growth) was input assuming that the shape of the crystal surface is flat, but the shape of the crystal surface may be input. When inputting the shape of the crystal surface, the configuration of the crystal cross-section may also be input. When inputting the configuration of the crystal cross-section, the configuration of the crystal cross-section of a polycrystalline crystal may be input, or the configuration of the crystal cross-section of a single crystal containing defects or impurities may be input.

[0097] In the embodiments described above, the details of the input method for the operator to input the crystal surface position (or crystal growth amount) as a manufacturing condition were not mentioned. However, the operator may estimate and input the crystal surface position (or crystal growth amount) based on the operating conditions.

[0098] In the embodiments described above, details of the information indicating the structure of the crystal growth furnace were not mentioned. However, the information indicating the structure of the crystal growth furnace may be, for example, CAD (Computer Aided Design) data. Alternatively, if deposits are present on the inner wall of the crystal growth furnace during operation, the CAD data may reflect the state in which such deposits are present.

[0099] In the embodiments described above, the temperature distribution of the crystal surface was shown in one dimension (see Figure 7), but the temperature distribution of the crystal surface may also be shown in two dimensions. The two-dimensional temperature distribution of the crystal surface refers to, for example, the temperature distribution within the plane of the circle when the crystal surface can be considered to be approximately a circle (for example, the difference in temperature at each position within the plane is represented by the difference in color).

[0100] In the second embodiment described above, the method for determining the various manufacturing conditions included in the training data was not mentioned. For example, the various manufacturing conditions may be determined by defining the range of each item included in the manufacturing conditions and then randomly determining the value of each item within the determined range.

[0101] In the second embodiment described above, the case in which the temperature field calculation device 410, the learning device 920, and the support device 910 are configured as separate devices was explained. However, the temperature field calculation device 410, the learning device 920, and the support device 910 may be configured as an integrated device. Alternatively, any two of the devices may be configured as an integrated device.

[0102] It should be noted that the present invention is not limited to the configurations shown in the above embodiments, including combinations with other elements. These aspects can be modified without departing from the spirit of the present invention and can be appropriately determined according to their application.

[0103] This application claims priority based on Japanese Patent Application No. 2024-204819, filed on 25 November 2024, which is incorporated herein by reference to the entire contents of the said Japanese Patent Application.

[0104] 10: Crystal manufacturing system 110: Crystal growth furnace 120: Control device 140: Support device 150: User interface device 410: Temperature field calculation device 610: Prerequisite setting unit 620: Manufacturing condition acquisition unit 630: Calculation unit 631: Temperature field calculation unit 640: Display unit 710, 720: Display screen 910: Support device 920: Learning device 1000: Learning data 1100: Learning unit 1310: Calculation unit 1311: Surrogate model

Claims

1. A support device comprising: a calculation unit that outputs the temperature distribution of the crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and a display unit that displays the temperature distribution of the crystal surface output by the calculation unit.

2. The support device according to claim 1, wherein if information indicating the position or output of the heating device is changed in response to the display of the temperature distribution on the crystal surface, the calculation unit recalculates based on the changed information, and the display unit updates the display of the temperature distribution on the crystal surface based on the recalculation result.

3. The support device according to claim 1 or 2, wherein the calculation unit has a physical model in which information indicating the structure of the crystal growth furnace, physical properties of each component, boundary conditions between each component, and information indicating the chemical structure of the raw material are set, and the physical model calculates the temperature field inside the crystal growth furnace when temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and the position of the crystal surface in the crystal growth furnace are input.

4. The support device according to any one of claims 1 to 3, wherein the calculation unit has a surrogate model trained using training data in which temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace are explanatory variables, and the temperature distribution of the crystal surface is the objective variable.

5. The support device according to any one of claims 1 to 4, wherein the display unit displays a display screen that includes: an input area for inputting temperature data measured in the crystal growth furnace, information indicating the position and output of a heating device for heating the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and an output area for displaying the temperature distribution of the crystal surface.

6. A computer support method comprising: a calculation step of calculating the temperature distribution of a crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and a display step of displaying the temperature distribution of the crystal surface calculated in the calculation step.

7. A support program for causing a computer to perform a calculation step of calculating the temperature distribution of the crystal surface based on temperature data measured in a crystal growth furnace, information indicating the position and output of a heating device that heats the crystal growth furnace, and information indicating the position of the crystal surface in the crystal growth furnace; and a display step of displaying the temperature distribution of the crystal surface calculated in the calculation step.

Images