Material analysis server, program used in material analysis server, and method used in material analysis server

Abstract

To reduce the burden of preprocessing of a system comprising multiple transmission devices for transmitting experiment result data, and a server for receiving and analyzing the experiment result data transmitted from the multiple transmission devices.SOLUTION: A material analysis server capable of receiving experiment data (12a) representing a result of an experiment on a material transmitted from multiple transmission devices (2) is provided, the material analysis server comprising a reception unit (13a) for receiving the experiment data (12a) transmitted by the multiple transmission devices, and a preprocessing unit (13b) configured to compute a feature quantity from a distribution of a physical quantity of the material included in each set of the experiment data and record a result it as preprocessed data (12b) for analysis.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a materials analysis server, a program used in the materials analysis server, and a method used in the materials analysis server. [Background technology]

[0002] Patent Document 1 describes a system that includes a transmitting device (e.g., an experimental device) that transmits data on experimental results related to materials, and a server that receives the data from the transmitting device. It also describes that in this system, the server performs analysis based on the experimental results (e.g., searching for test conditions that have optimal properties). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2021-193623 Summary of the Invention [Problem to be solved by the invention]

[0004] According to the inventor's investigations, in a system such as that of Patent Document 1, it is difficult for the server to directly use the experimental result data for analysis, and it is often necessary to perform preprocessing (e.g., noise removal, calculation of half-width) on the experimental result data for analysis.

[0005] Preprocessing work is difficult for material developers to perform alone and often requires the assistance of preprocessing experts. This is particularly true for preprocessing that involves calculating feature quantities from the distribution of physical quantities related to the material being tested. Therefore, if different experimental result data are sent from multiple transmission devices to a server and preprocessing is performed on each transmission device, it is highly likely that the work of multiple experts will be required. This means that the work of experts will be involved in the experiment-analysis cycle, which could reduce the efficiency of the cycle.

[0006] In view of the above, the present invention aims to reduce the burden of preprocessing in a system having multiple transmitting devices that transmit experimental result data and a server that receives and analyzes the experimental result data transmitted from the multiple transmitting devices. [Means for solving the problem]

[0007] To achieve the above object, the present invention provides a material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), the server comprising: a receiving unit (13a) for receiving the experimental data transmitted by the plurality of transmitting devices; a preprocessing unit (13b) that calculates feature quantities from the distribution of physical quantities related to the material contained in each of the experimental data and records the calculated feature quantities as preprocessed data (12b) for analysis. 、 Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data includes a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to the conditions of the experiment related to the signal data; the preprocessing unit calculates a feature from a body portion of each of the one or more signal data to generate the preprocessed data, and includes a part or all of a header portion of the signal data in a record corresponding to the signal data among the plurality of records, and records the resulting data as formed data (12c). This is a materials analysis server.

[0008] The invention described in claim 6 is a program used in a material analysis server that receives experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), the program comprising: a calculation unit (S200) for calculating feature quantities from the distribution of physical quantities related to materials included in each of the experimental data to obtain preprocessed data (12b); and A recording unit (S300) for recording the preprocessed data for analysis. 、S400 ) and make the material analysis server function as 、 Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data includes a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to the conditions of the experiment related to the signal data; the calculation unit calculates a feature from a body portion of each of the one or more signal data to generate the preprocessed data, and the recording unit includes a part or all of a header portion of the signal data in a record corresponding to the signal data among the plurality of records, and records the resulting data as formed data (12c). It is a program.

[0009] The invention described in claim 7 is a method performed by a material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), The material analysis server receiving the experimental data transmitted by the plurality of transmitting devices; The material analysis server Calculating feature quantities from the distribution of physical quantities related to the material contained in each of the experimental data, and recording the calculated feature quantities as preprocessed data (12b) for analysis. 、 Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data includes a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to the conditions of the experiment related to the signal data; In the calculating of the feature amount, the material analysis server calculates the feature amount from a body portion of each of the one or more signal data to obtain the preprocessed data, and in the recording, includes a part or all of a header portion of the signal data in a record corresponding to the signal data among the plurality of records, and records the resulting data as formed data (12c). It is a method.

[0010] In this way, by having the materials analysis server perform preprocessing on each of the experimental data transmitted from the multiple transmitting devices, the burden of preprocessing can be reduced compared to when preprocessing is performed individually on the transmitting device side. Furthermore, since this preprocessing is preprocessing that calculates feature quantities from the distribution of physical quantities related to the material, the effect of reducing the burden of preprocessing is more pronounced.

[0011] The reference symbols in parentheses attached to each component indicate an example of the correspondence between the component and the specific components described in the embodiments described below. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 is a configuration diagram of a material analysis system. [Figure 2] 1 is an example of a hardware configuration of a server. [Figure 3] FIG. 2 is a diagram illustrating a functional configuration of a server. [Figure 4] FIG. 1 is a diagram illustrating container virtualization. [Figure 5] FIG. 1 is a diagram showing the structure of experimental data. [Figure 6] 10 is a flowchart showing the processing contents of a preprocessing unit. DETAILED DESCRIPTION OF THE INVENTION

[0013] An embodiment of the present disclosure will be described below. As shown in Fig. 1, a material analysis system according to this embodiment includes a server 1, a plurality of transmitting devices 2, and a user terminal 3. The server 1, the transmitting devices 2, and the user terminal 3 are capable of communicating with each other via a communication network 4. The communication network 4 may include a wide area network such as the Internet. Furthermore, the communication network 4 may include a wired network or a wireless network.

[0014] Each of the multiple transmitting devices 2 conducts an experiment on a material based on the operation of an experimenter or a predetermined procedure, and records experimental data indicating the experimental results in a storage medium of the transmitting device 2. In other words, the transmitting device 2 is an experimental device.

[0015] The material to be tested may be different for each transmitting device 2, or may be the same. Furthermore, experiments on different materials may be conducted on the same transmitting device 2, or experiments on a single material may be conducted. Types of experiments include, for example, X-ray spectroscopy, X-ray diffraction measurement, infrared spectroscopy, X-ray imaging, infrared imaging, and microscope imaging.

[0016] Furthermore, the transmitting device 2 transmits the experimental data recorded on the storage medium to the server 1 via the communication network 4. The number of experimental results transmitted at one time from one transmitting device 2 as experimental data may be one or more.

[0017] Each of the transmitting devices 2 has experimental devices (e.g., X-ray irradiation devices, sensors, cameras, motors, etc.) for conducting experiments on materials, and each of the transmitting devices 2 also has a control circuit for controlling the operation of the experimental devices, a storage medium for storing experimental data, and a communication circuit for transmitting the experimental data to the server 1 via the communication network 4.

[0018] As another example, the transmitting device 2 itself may not have an experimental device, and experimental data resulting from an experiment conducted by a separately provided experimental device may be acquired, for example, by an operator performing a data transfer operation. The acquired experimental data may then be transmitted to the server 1 via the communication network 4.

[0019] The user terminal 3 is a terminal for obtaining information from the server 1. The user terminal 3 includes, for example, an operation unit that accepts user operations and converts them into electrical signals, a communication circuit that communicates with the server 1 via the communication network 4, a display that displays images to the user, and a processing circuit. The processing circuit generates request data based on the electrical signal corresponding to the user operation output from the operation unit and transmits it to the server 1 using the communication circuit. The control circuit then receives information returned from the server 1 in response to the request data using the communication circuit, and displays the received information on the display.

[0020] The server 1 is a material analysis server that receives experimental data transmitted from a plurality of transmitting devices 2 and performs analysis using the received experimental data. The server 1 then transmits the analysis results to the user terminal 3 in response to a request or the like transmitted from the user terminal 3 via the communication network 4. The user terminal 3 displays the received analysis results on a display for the operator of the user terminal 3 as described above.

[0021] 2, the server 1 has a communication interface circuit 11, a storage medium 12, and a control circuit 13. The communication interface circuit 11 is an interface circuit for communicating with the transmitting device 2 and the user terminal 3 via the communication network 4.

[0022] The storage medium 12 is a non-volatile storage medium such as an SSD or HDD. A program executed by the control circuit 13 is recorded in the storage medium 12. Note that the program in this embodiment may be configured in any form that defines the operation of the control circuit 13, such as source code, script, or binary. Furthermore, data can be read from and written to the storage medium 12 by the control circuit 13. The storage medium 12 is a non-transient tangible storage medium.

[0023] The control circuit 13 is a circuit having an arithmetic circuit, a volatile storage medium such as RAM, etc. The arithmetic circuit performs various processes by executing various programs recorded in the storage medium 12, and uses the volatile storage medium as a working area during this process. The volatile storage medium is a non-transient tangible storage medium.

[0024] 3, the control circuit 13 functions as a receiving unit 13a, a preprocessing unit 13b, and an analyzing unit 13c. Specifically, the control circuit 13 executes a receiving unit program, a preprocessing unit program, and an analyzing unit program recorded in the storage medium 12, thereby functioning as the receiving unit 13a, the preprocessing unit 13b, and the analyzing unit 13c, respectively. However, as another example, the control circuit 13 may have a dedicated circuit for the receiving unit 13a, a dedicated circuit for the preprocessing unit 13b, and a dedicated circuit for the analyzing unit 13c. In this case, these dedicated circuits may be hardware circuits whose circuit configuration is not programmable, or may be programmable logic circuits whose circuit configuration is programmable.

[0025] The receiving unit 13a operates continuously when the server 1 is in operation. Then, every time experimental data is transmitted to the server 1 from one of the above-mentioned multiple transmitting devices 2, the receiving unit 13a receives the experimental data via the communication interface circuit 11 and records the received experimental data 12a in an experimental data storage area of ​​the storage medium 12. Furthermore, every time new experimental data 12a is received and recorded in the storage medium 12, the receiving unit 13a uses this as a trigger to repeatedly activate the preprocessing unit 13b, as shown by the dashed arrow in Figure 3.

[0026] The preprocessing unit 13b is activated by the receiving unit 13a based on the trigger as described above. The preprocessing unit 13b then performs preprocessing, which is necessary for the experimental data 12a recorded by the receiving unit 13a as described above, as a pre-stage process prior to the analysis process by the analyzing unit 13c, which will be described later. The preprocessing creates preprocessed data 12b, formed data 12c, and summary data 12d based on the experimental data 12a, and records these created data in predetermined areas of the storage medium 12.

[0027] Here, the execution form of the preprocessing unit 13b will be described with reference to FIG. 4. The preprocessing unit 13b is a container executed in a container-type virtualized environment. Specifically, the control circuit 13 functions as a receiving unit 13a, a preprocessing unit 13b, and an analyzing unit 13c, and also functions as an OS unit 13y and a container runtime unit 13z. OS stands for Operating System. The control circuit 13 executes an OS program and a container runtime program recorded in the storage medium 12, thereby functioning as the OS unit 13y and the container runtime unit 13z, respectively.

[0028] The OS unit 13y operates as an OS, and performs file system management in the storage medium 12, management of processes running on the OS, control of the communication interface circuit 11, etc., while providing various interfaces to the processes running on the OS. The processing of the receiving unit 13a, the analyzing unit 13c, and the container runtime unit 13z is a process that runs on this OS.

[0029] The container runtime unit 13z runs on the OS and provides a container-type virtualized environment. Containers run using these container-type virtualized environments. In the example of Figure 4, the preprocessing unit 13b and the separate container 13p each correspond to a container.

[0030] For example, Docker may be adopted as the container runtime unit 13z, and a Docker container may be adopted as the container, but other combinations of container runtime and container may also be adopted.

[0031] The container corresponding to the preprocessing unit 13b has an application unit 131b and a resource unit 132b. The application unit 131b is a unit that realizes the main operation of the preprocessing unit 13b. For example, the application unit 131b corresponds to a part of the preprocessing program that selects data to be used in preprocessing from the experimental data 12a, selects the type of calculation to be performed on the selected data, and specifies the output format of the calculation results. By executing a binary executable file or an executable script file corresponding to these parts, the control circuit 13 functions as the application unit 131b.

[0032] The resource unit 132b functions as a resource used by the application unit 131b. For example, the resource unit 132b functions as a library called by the application unit 131b. Furthermore, if the application unit 131b is written in a script language such as Python, the resource unit 132b also functions as an interpreter for the script language. The resource unit 132b also defines a file system that is the file input / output destination for the application unit 131b.

[0033] The separate container 13p may perform any processing. The control circuit 13 functions as the separate container 13p by executing a program for the separate container recorded in the storage medium 12. The separate container 13p also has an application unit 131p and a resource unit 132p in addition to the pre-processing unit 13b. The application unit 131p is a unit that realizes the main operation of the separate container 13p. The resource unit 132p functions as a resource used by the application unit 131p.

[0034] The container runtime unit 13z manages the processing processes of the preprocessing unit 13b and the other container 13p and manages the memory within the preprocessing unit 13b and the other container 13p. These management operations are performed for each container in an environment independent of other containers. That is, the processing processes and memory resources of the preprocessing unit 13b are managed independently of the processing processes and memory of the other container 13p, in an isolated state protected from the processing of the other container 13p.

[0035] In this way, containers operate independently of each other. This reduces the possibility that the processing of another container 13p will interfere with the processing of the pre-processing unit 13b, and vice versa. For example, the resource unit 132b of the pre-processing unit 13b cannot be used by the application unit 131p of another container 13p.

[0036] In response to the request data received from the user terminal 3, the analysis unit 13c performs analysis using the preprocessed data 12b, the formed data 12c, and the summary data 12d recorded in the storage medium 12, and transmits information on the analysis results to the user terminal 3.

[0037] The operation of the material analysis system configured as described above will be described below. First, each of the transmitting devices 2 transmits experimental data to the server 1 via the communication network 4. The transmission timing may be independent for each transmitting device 2, or may be linked. Each of the transmitting devices 2 may transmit experimental data to the server 1 each time a user of the transmitting device 2 (e.g., an experimenter) performs a transmission operation (e.g., an operation to start a transmission batch process). Furthermore, each of the transmitting devices 2 may transmit one or more pieces of experimental data newly generated or acquired by the transmitting device 2 during each predetermined period to the server 1 in a lump.

[0038] In the server 1, each time a piece of experimental data arrives, the receiving unit 13a receives the experimental data 12a and records it in the storage medium 12. Each piece of experimental data 12a recorded in the storage medium 12 is raw data before being subjected to preprocessing.

[0039] 5, one piece of experimental data 12a includes one or more pieces of signal data 121 and one piece of list data 122. Each piece of signal data 121 corresponds to one experiment. Therefore, the number of pieces of signal data 121 included in the experimental data 12a corresponds to the number of experimental results included in the experimental data 12a.

[0040] Each piece of signal data 121 has a header section 121a and a body section 121b. The header section 121a includes information relating to various conditions of the experiment related to the signal data 121. For example, the header section 121a includes information such as the name of the measuring device, the name of the measurement date and time, the start and end points of the energy range of X-rays used in the measurement, and the name of the physical quantity represented in the body section 121b.

[0041] The body part 121b is a numerical sequence data representing a distribution of a physical quantity related to the material that was the subject of the experiment related to the signal data 121. The body part 121b may be, for example, an X-ray absorption spectrum obtained by X-ray spectrometry and representing an energy domain distribution of X-ray absorption coefficients. The body part 121b may also be an X-ray scattering spectrum obtained by X-ray diffraction measurement and representing an angular domain distribution of X-ray intensity. The body part 121b may also be an X-ray image obtained by X-ray imaging of the material surface and representing a two-dimensional spatial domain distribution of X-ray intensity. Note that the distribution may be any physical quantity domain distribution, such as a time domain distribution, in addition to the energy domain distribution, angular domain distribution, and spatial domain distribution described above.

[0042] The list data 122 has attribute information of the target signal data 121 for each of all signal data 121 included in the experimental data 12a to which the list data 122 belongs. Each piece of attribute information may have, for example, the file name of the target signal data 121, the name of the material used in the experiment related to the signal data 121, the experimenter, and other information. Therefore, the list data 122 has the same number of records as the signal data 121 included in the experimental data 12a to which the list data 122 belongs. There is a one-to-one correspondence between the signal data 121 included in one piece of experimental data 12a and the same number of records as the signal data. Each record has attribute information of the corresponding signal data 121. Each row of the list data 122 shown in FIG. 5 corresponds to one record.

[0043] The receiving unit 13a starts the preprocessing unit 13b at the timing when a trigger occurs, that is, when the experimental data 12a is received and recorded in the storage medium 12. Every time the preprocessing unit 13b is started, it executes the process shown in FIG.

[0044] Specifically, in step S100, the preprocessing unit 13b performs processing to separate each of the signal data 121 in the received experimental data 12a into a header portion 121a and a body portion 121b.

[0045] Next, in step S200, the preprocessing unit 13b performs various arithmetic operations on each body part 121b obtained as a result of the separation, thereby calculating preprocessed data 12b from each body part 121b. The preprocessed data 12b is data representing the feature quantities of the body part 121b from which the data was calculated. The preprocessed data 12b representing the feature quantities may be normalized data obtained by normalizing the spectrum of the body part 121b. The preprocessed data 12b representing the feature quantities may also be EXAFS amplitude data for the X-ray absorption spectrum represented by the body part 121b. EXAFS stands for Extended X-ray Absorption Fine Structure. The preprocessed data 12b representing the feature quantities may also be RDF. RDF stands for Radial Distribution Function. The preprocessed data 12b may include multiple types of data on the feature quantities listed above.

[0046] Next, the preprocessing unit 13b proceeds to step S300 and records the preprocessed data 12b calculated in step S200 in the storage medium 12.

[0047] Next, in step S400, the pre-processing unit 13b creates shaped data 12c and summary data 12d from list data 122 belonging to the same experimental data 12a as the signal data 121 and each header portion 121a obtained as a result of the separation, and records these in the storage medium 12.

[0048] Specifically, part or all of the contents of the header portion 121a of each signal data 121 is included in the record in the list data 122 that corresponds to that signal data 121. As a result, the list data 122 becomes the formed data 12c.

[0049] The summary data 12d thus created includes a portion of the list data 122 and a portion of each header section 121a. For example, the summary data 12d includes a brief description of each experiment. This description may be recorded in each record of the list data, or may be included in the experiment data 12a as data other than the signal data 121 and the list data 122.

[0050] After step S400, one processing run of the preprocessing unit 13b is completed. By repeatedly executing the preprocessing unit 13b in this manner, the preprocessed data 12b, the shaped data 12c, and the outline data 12d are accumulated in the storage medium 12.

[0051] Then, suppose that a user performs a predetermined operation on the user terminal 3 while a plurality of pieces of experimental data 12a, preprocessed data 12b based on the experimental data 12a, formed data 12c, and summary data 12d are stored in the storage medium 12. Then, the user terminal 3 transmits request data corresponding to the predetermined operation to the server 1 via the communication network 4.

[0052] The analysis unit 13c operates continuously while the server 1 is in operation. When the request data arrives at the server 1, the analysis unit 13c performs an analysis corresponding to the request data using the preprocessed data 12b, the formed data 12c, and the summary data 12d recorded in the storage medium 12. The analysis unit 13c then returns the analysis result to the user terminal 3 that is the sender of the request data.

[0053] The analysis performed may be any one of prediction, visualization, and machine learning, a combination of any two, or all three, or may be an analysis that is neither prediction, visualization, nor machine learning.

[0054] Visualization is a process of expressing information in the form of a graph, a scatter plot, a heat map, a contour, or other diagram using the preprocessed data 12b, the shaped data 12c, and the summary data 12d. For example, the analysis unit 13c may generate a two-dimensional graph with the absorption coefficient on the vertical axis and the energy on the horizontal axis for the EXAFS amplitude data recorded as the preprocessed data 12b, and transmit image data of the generated two-dimensional graph to the user via the communication network 4. Then, the user terminal 3 displays this image on its own display. The process of the analysis unit 13c at this time corresponds to visualization.

[0055] Furthermore, for example, the analysis unit 13c may perform calculations using a neural network in which the normalized data is used as an explanatory variable and the material name (e.g., chemical formula) of the material corresponding to the normalized data is used as a response variable. In this case, the analysis unit 13c performs training on the neural network using pairs of normalized data and material names derived from the same experimental data 12a obtained from a plurality of experimental data 12a as training data sets. The training may be performed, for example, every time the preprocessing unit 13b is executed before receiving request data.

[0056] In this case, upon receiving the request data, the analysis unit 13c inputs the normalized data included in the request data into the neural network using the trained neural network. The material name obtained as an output of the neural network is then transmitted to the user terminal 3 that sent the request data via the communication network 4. The processing by the analysis unit 13c at this time corresponds to machine learning, and corresponds to predicting the material name based on the normalized data. Note that in machine learning, a function approximator other than a neural network may be used.

[0057] Furthermore, the analysis unit 13c may also transmit, together with the analysis result, data from the formed data 12c and the summary data 12d that are related to the same experiment as the data used in calculating the analysis result, to the user terminal 3 that is the sender of the request data. In this case, the user terminal 3 may display the transmitted data on its own display.

[0058] As described above, the preprocessing unit 13b calculates feature quantities from the distribution of physical quantities related to the material included in each of the received experimental data 12a and that was the subject of the experiment related to the experimental data 12a, and sets the feature quantities as preprocessed data 12b, and records the preprocessed data 12b for analysis.

[0059] In this way, the server 1 performs preprocessing on each piece of experimental data transmitted from the multiple transmitting devices 2, thereby reducing the burden of preprocessing compared to when preprocessing is performed individually on the transmitting device 2 side. Furthermore, since this preprocessing is preprocessing for calculating feature quantities of the distribution of physical quantities related to the material, the effect of reducing the burden of preprocessing is more pronounced.

[0060] (1) Furthermore, the preprocessing unit 13b is a container executed in a container-type virtualized environment. Therefore, even if it is desired to change the processing content of the preprocessing unit 13b to something else, this can be achieved by simply replacing the current preprocessing unit 13b with a different preprocessing unit 13b along with the container on the server 1. In other words, simply replacing the container can accommodate a variety of analyses. An example of a case in which it is desired to change the processing content of the preprocessing unit 13b to something else is when the format or type of experimental data has changed.

[0061] Without a container-based virtualized environment, it may be difficult to replace the preprocessing unit 13b. For example, suppose that the programs that realize the application units 131b of two preprocessing units 13b are written in a scripting language such as Python. In addition, a library called by the application unit 131b of the first preprocessing unit 13b may differ from another library called by the application unit 131b of the second preprocessing unit 13b. Furthermore, the interpreter versions required by these two libraries may conflict with each other.

[0062] In such a case, without a container-type virtualized environment, it is highly likely that a complicated process would be required to interchange these two preprocessing units 13b. This is because it would be necessary to uninstall one library and its corresponding version of the interpreter, and then install the other library and its corresponding version of the interpreter. Container virtualization solves this problem. In the case of materials analysis, such container virtualization technology is particularly useful because there are a variety of data formats that must be handled.

[0063] Furthermore, two or more preprocessing units 13b with different processing contents may be provided in the server 1. In this case, these two or more preprocessing units 13b are realized as separate containers. This reduces the possibility that the resource unit 132b of one preprocessing unit 13b will adversely affect the function of the resource unit 132b of another preprocessing unit 13b.

[0064] Furthermore, the development of the preprocessing unit program may be performed in a different environment other than the server 1, for example, on a computer owned by a software developer. In such a case, the preprocessing unit program can be created as a container in the different environment by installing the same container runtime unit 13z as on the server 1. By using the preprocessing unit program created in this way on the server 1, the preprocessing unit 13b on the server 1 is more likely to function properly.

[0065] (2) Furthermore, the receiving unit 13a starts the execution of the preprocessing unit 13b based on a trigger related to the reception of the experimental data, which allows the preprocessing unit 13b to be executed efficiently since the preprocessing unit 13b is executed in association with the reception of the experimental data.

[0066] (3) Furthermore, the analysis unit 13c that analyzes the preprocessed data 12b realizes at least one of prediction based on the preprocessed data 12b, visualization using the preprocessed data 12b, and machine learning that includes the preprocessed data 12b in a learning dataset.

[0067] In this way, preprocessing is performed in the server 1 having the analysis unit 13c that realizes at least one of prediction, visualization, and machine learning. Therefore, by transmitting the experimental data to the server 1 that performs the preprocessing, it is also possible to provide the experimental data to be used for processing by the analysis unit 13c to the analysis unit 13c. In other words, it becomes easy to provide the experimental data to be used for processing by the analysis unit 13c to the analysis unit 13c.

[0068] (4) Furthermore, the preprocessing unit 13b calculates features from the body portion 121b of each piece of signal data 121 to generate preprocessed data 12b, and also includes part or all of the header portion 121a of the signal data 121 in a record in the list data 122 corresponding to the signal data 121, thereby recording the result as formed data 12c. In this way, the header portion 121a and the list data 122 can also be integrated on the server 1 side, which improves convenience for the experimenter.

[0069] In this embodiment, the control circuit 13 functions as a calculation unit by executing step S200 in Fig. 6. Also, it functions as a recording unit by executing step S300.

[0070] (Other embodiments) The present invention is not limited to the above-described embodiments and can be modified as appropriate. The above-described embodiments are not unrelated to each other and can be combined as appropriate unless the combination is clearly impossible. In the above-described embodiments, the elements constituting the embodiments are not necessarily essential unless expressly stated as essential or clearly considered essential in principle. In the above-described embodiments, when numerical values ​​such as the number, value, amount, and range of components of the embodiments are mentioned, they are not limited to the specific number unless expressly stated as essential or clearly limited to a specific number in principle. In particular, when multiple values ​​are exemplified for a certain quantity, values ​​between those multiple values ​​can be adopted unless otherwise specified or clearly impossible in principle. In the above-described embodiments, when the shape, positional relationship, etc. of components are mentioned, they are not limited to the shape, positional relationship, etc., unless expressly stated or limited to a specific shape, positional relationship, etc. in principle. The present invention also allows the following modifications and modifications within equivalent ranges to the above-described embodiments. The following modifications can be independently applied or inapplicable to the above-described embodiments. That is, any combination of the modified examples described below can be applied to the above embodiment.

[0071] The control circuit 13 and techniques described herein may also be implemented by a special-purpose computer configured with a processor and memory programmed to perform one or more functions embodied in a computer program. Alternatively, the control circuit 13 and techniques described herein may be implemented by a special-purpose computer configured with one or more dedicated hardware logic circuits. Alternatively, the control circuit 13 and techniques described herein may be implemented by one or more special-purpose computers configured with a processor and memory programmed to perform one or more functions in combination with a processor configured with one or more hardware logic circuits. Furthermore, the computer program may be stored as instructions executed by a computer on a computer-readable non-transitory storage medium.

[0072] (Variation 1) In the above embodiment, the preprocessing performed by the preprocessing unit 13b includes, for example, spectrum normalization, calculation of EXAFS amplitude data, and calculation of RDF. However, the preprocessing performed by the preprocessing unit 13b is not limited to these. For example, the preprocessing performed by the preprocessing unit 13b may include noise removal, smoothing, and trimming of the spectrum. The data after noise removal, smoothing, and trimming are also spectral feature quantities.

[0073] Furthermore, for example, if the experimental data 12a is obtained by infrared spectroscopy, the preprocessing performed by the preprocessing unit 13b may be calculation of the peak top, half width, area, etc. of the IR spectrum. Furthermore, if each signal data 121 in the experimental data 12a includes an image such as a microscope image, the preprocessing performed by the preprocessing unit 13b may be calculation of the proportion of an area occupied by a specific substance, the outline, etc.

[0074] (Variation 2) In the above embodiment, experimental data 12a showing the experimental results is exemplified as experimental data for real materials. However, the experimental data 12a showing the experimental results may be data showing the results of numerical experiments. An example of a numerical experiment on materials is a first-principles calculation that calculates energy distribution from the molecular structure.

[0075] (Variation 3) In the above embodiment, the reception of the experimental data 12a is exemplified as a trigger for the reception unit 13a to activate the pre-processing unit 13b. However, the trigger for the reception unit 13a to activate the pre-processing unit 13b may be another trigger related to the reception of the experimental data 12a.

[0076] For example, the trigger may be when the number of unpreprocessed experimental data 12a received by the receiving unit 13a reaches a predetermined number or more. Alternatively, the trigger may be when a predetermined period of time has elapsed since the last preprocessed experimental data 12a was received by the receiving unit 13a.

[0077] Alternatively, the trigger may be, for example, when the unpreprocessed experimental data 12a received by the receiving unit 13a includes the first type of experimental data 12a and the second type of experimental data 12a. In this case, the first type of experimental data 12a and the second type of experimental data 12a may be of different types of materials as experimental subjects. Alternatively, the first type of experimental data 12a and the second type of experimental data 12a may be of different experimental forms.

[0078] (Variation 4) In the above embodiment, the receiving unit 13a, the preprocessing unit 13b, and the analyzing unit 13c are all realized in the same server 1. However, this is not necessarily the case. For example, the receiving unit 13a and the preprocessing unit 13b may be realized in the server 1, and the analyzing unit 13c may be realized in a device other than the server 1. In this case, the other device is capable of communicating with the server 1 and the user terminal 3.

[0079] (Variation 5) In the above embodiment, the server 1 is exemplified as a single device, but the server 1 may be realized by a plurality of devices. That is, the server 1 may be realized as a cloud. In this case, data exchange between these multiple devices may be realized by communication via the communication network 4.

[0080] (Variation 6) In the above embodiment, the preprocessing unit 13b creates the formed data 12c from the list data 122 and the header portion 121a. However, the preprocessing unit 13b may create the formed data 12c from only the header portion 121a of the list data 122 and the header portion 121a. For example, this may be done when the list data 122 does not exist, or when the list data 122 exists.

[0081] Furthermore, the preprocessing unit 13b may include data created from the body unit 121b in addition to the list data 122 and the header unit 121a in the formed data 12c. For example, the preprocessing unit 13b may calculate representative values ​​such as the average value, variance, maximum value, and minimum value of the numeric string in the body unit 121b and include the calculation results in the formed data 12c. Furthermore, the preprocessing unit 13b may not create the formed data 12c from the list data 122 and the header unit 121a, but may include only the data created from the body unit 121b in the formed data 12c.

[0082] The preprocessing unit 13b may include in the formed data 12c data created from a portion of the signal data 121 that is neither the header portion 121a nor the body portion 121b. For example, the formed data 12c may include information on the file name of the signal data 121. Furthermore, the preprocessing unit 13b may not create the formed data 12c from the list data 122, the header portion 121a, or the body portion 121b, but may include in the formed data 12c only data created from a portion of the signal data 121 that is neither the header portion 121a nor the body portion 121b.

[0083] In this way, the preprocessing unit 13b records, for each piece of experimental data 12a, data created based on the experimental data 12a as shaped data 12c separately from the preprocessed data 12b. In this way, data used in a different usage pattern from that of the preprocessed data 12b can be managed separately as shaped data 12c. As yet another example, the preprocessing unit 13b may create the shaped data 12c only from the list data 122 without using the signal data 121.

[0084] (Features of the invention) [Claim 1] A material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), a receiving unit (13a) that receives the experimental data (12a) transmitted from the plurality of transmitting devices; a preprocessing unit (13b) that calculates feature quantities from the distribution of physical quantities related to the material contained in each of the experimental data and records the calculated feature quantities as preprocessed data (12b) for analysis. [Claim 2] 2. The materials analysis server according to claim 1, wherein the preprocessing unit is a container that runs in a container-type virtualized environment. [Claim 3] 3. The materials analysis server according to claim 1, wherein the receiving unit starts execution of the preprocessing unit based on a trigger related to reception of the experimental data. [Claim 4] an analysis unit (13c) that analyzes the preprocessed data; 4. The materials analysis server according to claim 1, wherein the analysis unit performs at least one of prediction based on the preprocessed data, visualization using the preprocessed data, and machine learning that includes the preprocessed data in a training dataset. [Claim 5] 5. The material analysis server according to claim 1, wherein the preprocessing unit records, for each of the experimental data, data created based on the experimental data as formed data (12c) separately from the preprocessed data. [Claim 6] Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data includes a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to the conditions of the experiment related to the signal data; 5. The material analysis server according to claim 1, wherein the preprocessing unit calculates features from a body portion of each of the one or more signal data to obtain the preprocessed data, and includes a part or all of a header portion of the signal data in a record corresponding to the signal data among the plurality of records, and records the resulting data as formed data (12c). [Claim 7] A program used in a material analysis server that receives experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), a calculation unit (S200) for calculating feature quantities from the distribution of physical quantities related to materials included in each of the experimental data to obtain preprocessed data (12b); and a program that causes the material analysis server to function as a recording unit (S300) that records the preprocessed data for analysis; [Claim 8] A method performed by a material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), comprising: receiving the experimental data transmitted by the plurality of transmitting devices; calculating feature quantities from the distribution of physical quantities related to the material contained in each of the experimental data and recording the result as preprocessed data (12b) for analysis. [Explanation of symbols]

[0085] 1 server 2. Transmitting device 12a Experimental data 12b Preprocessed data 13a Receiving section 13b Pretreatment section 13c Analysis Department

Claims

1. A material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), a receiving unit (13a) that receives the experimental data (12a) transmitted from the plurality of transmitting devices; a preprocessing unit (13b) that calculates feature quantities from the distribution of physical quantities related to materials included in each of the experimental data and records the feature quantities as preprocessed data (12b) for analysis; Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data comprises a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to conditions of the experiment related to the signal data; The preprocessing unit calculates features from the body of each of the one or more signal data to generate the preprocessed data, and includes part or all of the header of the signal data in a record among the plurality of records corresponding to the signal data, and records the resulting data as formed data (12c).

2. The materials analysis server according to claim 1 , wherein the preprocessing unit is a container that runs in a container-type virtualized environment.

3. 3. The material analysis server according to claim 1, wherein the receiving unit starts the execution of the preprocessing unit based on a trigger related to the reception of the experimental data.

4. an analysis unit (13c) that analyzes the preprocessed data, 3. The materials analysis server according to claim 1, wherein the analysis unit performs at least one of prediction based on the preprocessed data, visualization using the preprocessed data, and machine learning that includes the preprocessed data in a training dataset.

5. 3. The material analysis server according to claim 1, wherein the preprocessing unit records, for each of the experimental data, data created based on the experimental data as formed data (12c) separately from the preprocessed data.

6. A program used in a material analysis server that receives experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), comprising: a calculation unit (S200) for calculating feature quantities from the distribution of physical quantities related to materials included in each of the experimental data to obtain preprocessed data (12b); and causing the material analysis server to function as a recording unit (S300, S400) that records the preprocessed data for analysis; Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data comprises a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to conditions of the experiment related to the signal data; The calculation unit calculates features from the body part of each of the one or more signal data to generate the preprocessed data, and the recording unit includes part or all of the header part of the signal data in a record among the plurality of records corresponding to the signal data, and records the resulting data as shaped data (12c).

7. A method performed by a material analysis server capable of receiving experimental data (12a) indicating experimental results on materials transmitted from a plurality of transmitting devices (2), comprising: the material analysis server receiving the experimental data transmitted by the plurality of transmitting devices; the material analysis server calculates feature quantities from distributions of physical quantities related to the materials included in each of the experimental data, and records the feature quantities as preprocessed data (12b) for analysis; Each of the experimental data includes one or more signal data (121) each representing a single experimental result, and list data (122) including one or more records corresponding to the one or more signal data; Each of the one or more signal data comprises a body portion (121b) representing a distribution of a physical quantity related to a material that was the subject of an experiment related to the signal data, and a header portion (121a) including information related to conditions of the experiment related to the signal data; the material analysis server, in calculating the feature quantity, calculates the feature quantity from a body portion of each of the one or more signal data to obtain the preprocessed data, and, in recording, includes part or all of a header portion of the signal data in a record among the plurality of records corresponding to the signal data, and records the resulting data as formed data (12c).

Images