Information processing device, inference device, machine learning device, information processing method, inference method, and machine learning method

Abstract

To provide an information processing device capable of acquiring optimal software setting information conforming to simulation specification information as an analysis target.SOLUTION: An information processing device 6 comprises: an information acquisition unit 600 that acquires simulation specification information as an analysis target; and a generation processing unit 601 that, by inputting the simulation specification information acquired by the information acquisition unit 600 to a learning model 12 that has learned by machine learning a correlation between the simulation specification information as the analysis target and software setting information for use when performing a simulation in accordance with the simulation specification information as the analysis target, generates the software setting information for the simulation specification information as the analysis target.SELECTED DRAWING: Figure 9

Description

[Technical field]

[0001] The present invention relates to an information processing device, an inference device, a machine learning device, an information processing method, an inference method, and a machine learning method. [Background technology]

[0002] For example, simulation of the flow behavior of a fluid is mainly performed using lattice-based CFD (Computational Fluid Dynamics) calculation software, which divides a space into a lattice and performs calculations (e.g., Patent Document 1). Currently, there are many commercially available software programs based on the lattice-based CFD, and these programs are widely used. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2021-144626 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the case of similar analysis software, including the simulation software for flow behavior analysis as described above, the amount of manuals and tutorials provided to users is enormous.

[0005] In order to properly run a simulation using software, it is extremely difficult to understand the simulation specifications of the target to be analyzed, and also to understand the provided manuals, tutorials, and, if necessary, references in the manual, in order to select which model to use, what kind of grid to use, what numerical calculation method to use, and what kind of display to display, all of which are provided by the software.

[0006] Making the above selections that match the simulation specifications of the target to be analyzed appropriately has a large impact on the success of the simulation, so the setting information for the simulation software should be selected by an expert with in-depth knowledge of the software, but in reality this is not the case.

[0007] In consideration of the above problems, the present invention aims to provide an information processing device, an inference device, a machine learning device, an information processing method, an inference method, and a machine learning method that enable easy prediction of software setting information related to what settings are optimal when performing a simulation in accordance with simulation specification information of the object to be analyzed using simulation software for fluids, etc. [Means for solving the problem]

[0008] In order to achieve the above object, an information processing device according to one aspect of the present invention comprises: an information acquisition unit for acquiring simulation specification information of an analysis target; The simulation specification information acquired by the information acquisition unit is input into a learning model that has been trained by machine learning to determine the correlation between the simulation specification information of the analysis target and software setting information when performing a simulation according to the simulation specification information of the analysis target, thereby generating the software setting information for the simulation specification information of the analysis target. Effect of the Invention

[0009] According to an information processing device of one embodiment of the present invention, by setting up a simulation based on software setting information generated by inputting simulation specification information of the analysis target into a learning model, it is possible to perform an optimal simulation that is in line with the simulation specification information of the analysis target.

[0010] Other objects, configurations and effects will become apparent from the detailed description of the invention described below. [Brief description of the drawings]

[0011] [Figure 1] 1 is an overall configuration diagram showing an example of a software operation support system 1. FIG. [Diagram 2] FIG. 2 is a diagram for explaining a processing image in the natural language processing unit 17. [Diagram 3] 1 is a diagram for explaining an image of how learning data 13 is created by a natural language processing unit 17. FIG. [Figure 4] FIG. 9 is a hardware configuration diagram showing an example of a computer 900. [Diagram 5] FIG. 2 is a block diagram showing an example of a machine learning device 5 according to an embodiment of the present invention. [Figure 6] FIG. 2 is a diagram showing an example of a learning model 12 and learning data 13. [Figure 7] 4 is a flowchart showing an example of a machine learning method performed by the machine learning device 5. [Figure 8] FIG. 2 is a block diagram showing an example of an information processing device 6 according to an embodiment of the present invention. [Figure 9] 1 is a functional explanatory diagram showing an example of an information processing device 6 according to an embodiment of the present invention. [Figure 10] 13 is a flowchart showing an example of an information processing method by the information processing device 6. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment for carrying out the present invention will be described with reference to the drawings. In the following, the scope necessary for the explanation to achieve the object of the present invention will be shown in a schematic manner, and the scope necessary for the explanation of the relevant part of the present invention will be mainly explained, and the parts that are omitted from the explanation will be based on the publicly known technology.

[0013] 1 is an overall configuration diagram showing an example of a software operation support system 1. When simulation specification information of an analysis target is given in simulation software used for analyzing the flow behavior of a fluid, the software operation support system 1 supports a user on how to set up the simulation software in order to execute the simulation using the simulation software.

[0014] The specific support by the software operation support system 1 is to provide the user with software setting information (item names of information related to which model to select, which grid to use, which numerical calculation method to use, what output display to perform, etc., in the simulation software, or in some cases, quantified information) when performing a simulation according to simulation specification information of the analysis target (item names of specific information related to what kind of simulation to perform, or in some cases, quantified information). For this purpose, the software operation support system 1 has a function of inferring and outputting the corresponding software setting information when the simulation specification information of the analysis target is input. One of the methods that can be adopted for such inference is machine learning.

[0015] In order to provide the above-mentioned user support, it is necessary to fully understand the contents of the huge amount of manuals and tutorials provided with the simulation software, and then to select various software setting information that matches the simulation specification information. For this purpose, the software operation support system 1 uses a language processing terminal device 2. This language processing terminal device 2 performs natural language processing on data relating to the contents of the manual and tutorial descriptions (each of which is referred to as manual document data and tutorial document data), thereby extracting terms that may be included in simulation specification information, as well as keywords and parameters for setting up the simulation software.

[0016] As shown in Fig. 1, the software operation support system 1 may include, as its main components, a language processing terminal device 2, an administrator terminal device 3, a software operation support management device 4, a machine learning device 5, and an information processing device 6. Each of the devices 2 to 6 is, for example, configured as a general-purpose or dedicated computer (see Fig. 4 described later), and is connected to a wired or wireless network 7 so as to be able to transmit and receive various data to and from each other. Note that the number of the devices 2 to 6 and the connection configuration of the network 7 are not limited to the example in Fig. 1, and may be changed as appropriate.

[0017] The language processing terminal device 2 can be used as a terminal device to which manual document data and tutorial document data of a target simulation software are input and which analyzes each document data. The language processing terminal device 2 may be a stationary device or a portable device. The language processing terminal device 2 includes a natural language processing unit 17, which extracts terms, keywords, and parameters used for inference and machine learning from the manual document data and tutorial document data using a method such as text mining.

[0018] The language processing terminal device 2 can also be used as a terminal device that acquires various data such as document data from an information storage / information providing means (not shown) such as a server provided on a network such as the Internet or an intranet. When using the language processing terminal device 2 for such purposes, it is preferable that the language processing terminal device 2 is set so that an application program for acquiring data from the information storage / information providing means runs on the language processing terminal device 2.

[0019] The administrator terminal device 3 is a terminal device used by an administrator who manages the entire software operation support system 1, and may be a stationary device or a portable device. The administrator terminal device 3 accepts various input operations via a display screen or a reader / writer of, for example, an application program, a web browser, or the like, and displays various information via the display screen. The administrator terminal device 3 can be used as a terminal device for the administrator to set basic management items, such as which simulation software out of multiple existing simulation software is to be targeted for operation support in the system. The administrator terminal device 3 can be used to set what manual document data and tutorial document data are to be acquired, and to display various information for the administrator to refer to.

[0020] The software operation support management device 4 includes a software operation support management database 41 for managing data (manual document data, tutorial document data, etc.) stored corresponding to each of a plurality of existing simulation software. The software operation support management database 41 can also store data related to terms, keywords, parameters, etc., corresponding to each simulation software, which are acquired from the language processing terminal device 2.

[0021] The machine learning device 5 is a device that operates as a subject of the learning phase of machine learning. The machine learning device 5 acquires various information such as terms, keywords, and parameters registered in the software operation support management database 41 of the software operation support management device 4 as learning data 13, and generates a learning model 12 used in the information processing device 6 by machine learning based on the learning data 13. The trained learning model 12 is provided to the information processing device 6 via the network 7, a recording medium, or the like.

[0022] The information processing device 6 is a device that operates as a main body of the inference phase of machine learning. The management device 6 predicts software setting information for performing a simulation based on the simulation specification information of the analysis target using the learning model 12 generated by the machine learning device 5. The software setting information as the prediction result is provided to the manager terminal device 3, for example, and presented to an information provider such as a business operator or an administrator. The software setting information as the prediction result is also provided to the software operation support management device 4 and accumulated in the software operation support management database 41.

[0023] Next, a description will be given of a process for extracting terms that may be included in simulation specification information, keywords and parameters for setting simulation software, etc., which can be subjected to machine learning, by the natural language processing unit 17 in the language processing terminal device 2. Fig. 2 is a diagram for explaining a processing image in the natural language processing unit 17.

[0024] Manual document data is input to the natural language processing unit 17 of the language processing terminal device 2. The natural language processing unit 17 extracts terms that may be included in the simulation specification information to be analyzed from the input manual document data using a method such as text mining. The natural language processing unit 17 also extracts software setting information for performing a simulation according to the simulation specification information to be analyzed from the input manual document data.

[0025] Term information that may be included in the simulation specification information of the analysis target may include information related to the space of the analysis target and information related to the medium filled in the space. Information related to the space of the analysis target includes information such as the material of the space, the shape of the space, and the dimensions of the space. Information related to the medium of the analysis target includes information such as the name of the medium (compound name), the physical properties of the medium, and the temperature of the medium.

[0026] The keywords and parameter information contained in the software setting information when performing a simulation according to the simulation specification information of the analysis target includes model setting information, condition setting information, output method setting information, and the like.

[0027] The model setting information is information about a model that can be selected and set by the simulation software. In the figure, it is shown that the simulation software can select and set A model, B model, C model, etc. The condition setting information is information about the grid conditions, boundary conditions, application conditions, etc. that can be selected and set by the simulation software. The condition setting information also includes parameters such as grid interval, numerical range, and setting step interval. The output method setting information is information related to the display output that can be selected and set by the simulation software. The output method setting information includes color-coded output, vector display output, numerical display output, etc. The output method setting information may also include graph output, table output, vector diagram output, streamline output, contour line output, isosurface output, etc.

[0028] Each term included in the "simulation specification information" of the analysis target extracted by the natural language processing unit 17 can be used as information related to the input layer when performing machine learning. Also, keywords, parameters, etc. included in the "software setting information" extracted by the natural language processing unit 17 when performing a simulation according to the simulation specification information of the analysis target can be used as information related to the output layer when performing machine learning.

[0029] Next, the process of creating the learning data 13 by the natural language processing unit 17 as described above will be described. In the following, an example of performing an optimal simulation based on information obtained from the tutorial document data will be described as an example of creating the learning data 13. For example, the learning data 13 may be simulation examples created by an experienced user of the simulation software.

[0030] 3 is a diagram for explaining an image of creating the learning data 13 by the natural language processing unit 17. Tutorial document data is input to the natural language processing unit 17 of the language processing terminal device 2. The natural language processing unit 17 extracts, from each simulation case using a method such as text mining, "simulation specification information" to be analyzed and "software setting information" for performing the corresponding simulation from the input tutorial document data.

[0031] In the example shown in the figure, the simulation specification information extracted includes information such as that the space to be analyzed is a 30φ cylindrical pipe, that the medium is ●● oil with a specified viscosity, and that the temperature is 25°C. In addition, the software setting information extracted for performing such a simulation includes information such as that the model setting information indicates that model B is set, that the condition setting information indicates that a grid with 5 mm intervals is set, that the metal boundary condition is set, and that the application condition without disturbance is set, and that the output method setting information indicates that color-coded output is set.

[0032] Next, the computers constituting the language processing terminal device 2, the manager terminal device 3, the software operation support management device 4, the machine learning device 5, and the information processing device 6 will be described.

[0033] (Hardware configuration of each device) 4 is a hardware configuration diagram showing an example of a computer 900. Each of the devices 2 to 6 of the software operation support system 1 is configured by a general-purpose or dedicated computer 900.

[0034] 4, the computer 900 includes, as its main components, a bus 910, a processor 912, a memory 914, an input device 916, an output device 917, a display device 918, a storage device 920, a communication I / F (interface) unit 922, an external device I / F unit 924, an I / O (input / output) device I / F unit 926, and a media input / output unit 928. Note that the above components may be omitted as appropriate depending on the application of the computer 900.

[0035] The processor 912 is composed of one or more arithmetic processing devices (such as a central processing unit (CPU), a micro-processing unit (MPU), a digital signal processor (DSP), or a graphics processing unit (GPU)) and operates as a control unit that controls the entire computer 900. The memory 914 stores various data and programs 930, and is composed of, for example, a volatile memory (DRAM, SRAM, etc.) that functions as a main memory, a non-volatile memory (ROM), a flash memory, etc.

[0036] The input device 916 is composed of, for example, a keyboard, a mouse, a numeric keypad, an electronic pen, etc., and functions as an input unit. For example, a stylus or a data glove can be used as the input device 916. Furthermore, the input device 916 may be a device in which an electronic file, so-called a script file, in which various input commands are described, is stored in a storage device and read into software. The input device 916 allows a user to input various settings (numerical values, options, or position information by a pointer) using a keyboard and a mouse. The output device 917 is composed of, for example, a sound (audio) output device, a vibration device, etc., and functions as an output unit. Furthermore, the output device 917 may output a three-dimensional shape as a 3D printer, or may output to a storage device as a data file in which numerical values ​​are regularly arranged.

[0037] The display device 918 is composed of, for example, a liquid crystal display, an organic electroluminescence display, an electronic paper, a projector, or the like, and functions as an output unit. The display device 918 typically displays numerical values, graphs, tables, vector diagrams, flow lines, contour lines, isosurfaces, or the like on a display. The input device 916 and the display device 918 may be integrally configured, such as a touch panel display. The display device 918 may also include a head-mounted display or smart glasses, which are a type of display. The storage device 920 is composed of, for example, a hard disk drive (HDD), a solid state drive (SSD), or the like, and functions as a storage unit. The storage device 920 stores various data required for the execution of the operating system and the program 930.

[0038] In the present invention, a function that can perform input / output via at least an electronic file in a storage device, or input / output using a keyboard, mouse, and display can be adopted. A robot arm-like device that replaces a human hand can be used for keyboard and mouse operation, and input via the keyboard and mouse can be performed using software and dedicated hardware (such as a board that virtually emulates keyboard and mouse input). In addition, output from a display can be observed with a camera-like device, or even if there is no physical display, dedicated hardware can be used to directly read output from a graphics board and grasp the information.

[0039] For the simulation software, the mouse and keyboard are input devices 916 and the display is a display device 918; however, when the system of the present invention automatically executes the simulation software, the mouse and keyboard can be considered as output devices for operating the simulation software, and the display as a device for inputting information.

[0040] The communication I / F unit 922 is connected to a network 940 (which may be the same as the network 7 in FIG. 1) such as the Internet or an intranet by wire or wirelessly, and functions as a communication unit that transmits and receives data to and from other computers according to a predetermined communication standard. The external device I / F unit 924 is connected to an external device 950 such as a camera, a printer, a scanner, a reader / writer by wire or wirelessly, and functions as a communication unit that transmits and receives data to and from the external device 950 according to a predetermined communication standard. The I / O device I / F unit 926 is connected to an I / O device 960 such as various sensors and actuators, and functions as a communication unit that transmits and receives various signals and data, such as detection signals from sensors and control signals to actuators, between the I / O device 960. The media input / output unit 928 is composed of a drive device such as a DVD (Digital Versatile Disc) drive or a CD (Compact Disc) drive, a memory card slot, and a USB connector, and reads and writes data from and to a medium (non-temporary storage medium) 970 such as a DVD, a CD, a memory card, or a USB memory.

[0041] In the computer 900 having the above configuration, the processor 912 loads a program 930 stored in the storage device 920 into the memory 914, executes the program, and controls each unit of the computer 900 via the bus 910. The program 930 may be stored in the memory 914 instead of the storage device 920. The program 930 may be recorded in the medium 970 in an installable file format or an executable file format, and provided to the computer 900 via the media input / output unit 928. The program 930 may be provided to the computer 900 by downloading it via the network 940 through the communication I / F unit 922. In addition, the computer 900 may implement various functions, which are realized by the processor 912 executing the program 930, using, for example, an FPGA, an ASIC, an FPGA, an ASIC, or the like. It may be realized by hardware such as SIC.

[0042] The computer 900 is, for example, a stationary computer or a portable computer, and is an electronic device of any form. The computer 900 may be a client computer, a server computer, or a cloud computer. The computer 900 may also be applied to devices other than the devices 2 to 6. In addition, when the computer 900 is a portable computer, it is preferable that the computer 900 has a function of acquiring location information of its own location and is configured to provide the location information for various data processing.

[0043] (Machine Learning Device 5) 5 is a block diagram showing an example of a machine learning device 5 according to an embodiment of the present invention. The machine learning device 5 includes a control unit 50, a communication unit 51, a learning data storage unit 52, and a trained model storage unit 53.

[0044] The control unit 50 functions as a learning data acquisition unit 500 and a machine learning unit 501. The communication unit 51 is connected to an external device via the network 7, and functions as a communication interface for transmitting and receiving various types of data.

[0045] The learning data acquisition unit 500 is connected to external devices (e.g., the language processing terminal device 2, the administrator terminal device 3, the software operation support management device 4, etc.) via the communication unit 51 and the network 7, and acquires learning data 13 consisting of simulation specification information of the analysis target as input data and software setting information for performing the simulation as output data. The learning data 13 is data used as teacher data (training data), verification data, and test data in supervised learning. In addition, the software setting information is data used as a correct answer label in supervised learning.

[0046] The learning data storage unit 52 is a database that stores a plurality of sets of learning data 13 acquired by the learning data acquisition unit 500. The specific configuration of the database that constitutes the learning data storage unit 52 may be designed as appropriate.

[0047] The machine learning unit 501 performs machine learning using a plurality of sets of learning data 13 stored in the learning data storage unit 52. That is, the machine learning unit 501 inputs a plurality of sets of learning data 13 to the learning model 12, and generates a trained learning model 12 by having the learning model 12 learn a correlation between the simulation specification information and the software setting information included in the learning data 13. When performing machine learning, the machine learning unit 501 can employ any method, such as online learning, batch learning, and mini-batch learning. Note that the machine learning unit 501 may perform a predetermined pre-processing on the input data (simulation specification information) input to the learning model 12, or may perform a predetermined post-processing on the output data (software setting information) output from the learning model 12.

[0048] The trained model storage unit 53 is a database that stores the trained learning model 12 (specifically, a group of adjusted weight parameters) generated by the machine learning unit 501. The trained learning model 12 stored in the trained model storage unit 53 is provided to an actual system (for example, an information processing device 6) via a network 7, a recording medium, or the like. Note that, although the training data storage unit 52 and the trained model storage unit 53 are shown as separate storage units in FIG. 5, they may be configured as a single storage unit.

[0049] FIG. 6 is a diagram showing an example of the learning model 12 and the learning data 13 in the case where the simulation specification information includes information related to the space and medium of the analysis target. The learning data 13 used for the machine learning of the learning model 12 includes the simulation specification information and software setting information. It consists of:

[0050] The simulation specification information constituting the learning data 13 includes at least one of information related to the space to be analyzed and information related to the medium to be analyzed. The information related to the space to be analyzed may include information related to the material, shape, dimensions, etc. The information related to the medium to be analyzed may include information related to the name (compound name), physical properties, temperature, etc.

[0051] The software setting information constituting the learning data 13 is information related to the setting of software when simulating an analysis target, and may include at least one of model setting information, condition setting information, and output method setting information. For example, taking the model setting information as an example, the rate at which it is appropriate to set model A, the rate at which it is appropriate to set model B, the rate at which it is appropriate to set model C, etc. may be expressed as a stepped value or a continuous value. In the case of a continuous value, the value may be normalized to a predetermined range (for example, 0 to 1). Of the rates of models A to C, the model with the highest rate calculated can be set as the model recommended for setting.

[0052] The software setting information may indicate the settings of the software when simulating the analysis target, and any of the model setting information, condition setting information, and output method setting information may be omitted or may include other information. The definition of the software setting information may be changed as appropriate, in which case the data configuration of the learning model 12 and the output data in the learning data 13 may be changed as appropriate.

[0053] The learning data acquisition unit 500 acquires the learning data 13 by referring to various information registered in the software operation support management database 41 and by accepting input operations by the administrator from the administrator terminal device 3.

[0054] The learning model 12 employs, for example, a neural network structure, and includes an input layer 120, an intermediate layer 121, and an output layer 122. Synapses (not shown) that connect each neuron are laid between each layer, and each synapse is associated with a weight. A weight parameter group consisting of the weights of each synapse is adjusted by machine learning.

[0055] The input layer 120 has neurons in a number corresponding to the simulation specification information as input data, and each value of the simulation specification information is input to each neuron. The output layer 122 has neurons in a number corresponding to the software setting information as output data, and a prediction result (inference result) of the software setting information for the simulation specification information to be analyzed is output as output data. When the learning model 12 is configured as a regression model, the software setting information is output as a numerical value normalized to a predetermined range (e.g., 0 to 1). When the learning model 12 is configured as a classification model, the software setting information is output as a score (accuracy) for each class as a numerical value normalized to a predetermined range (e.g., 0 to 1).

[0056] The machine learning unit 501 inputs multiple sets of learning data 13 into the learning model 12, and generates a trained learning model 12 by having the learning model 12 learn the correlation between the simulation specification information contained in the learning data 13 and the software setting information.

[0057] In this embodiment, the data configuration of the learning model 12 and the learning data 13 is described as shown in FIG. 6. However, for example, a plurality of data with different conditions may be used depending on the machine learning method, the type of simulation software to be used, the type of data included in the simulation software, the type of data included in the software setting information, etc. In that case, the learning data acquisition unit 500 may acquire multiple types of learning data corresponding to multiple data configurations with different conditions, and the machine learning unit 501 may perform machine learning using each of the learning data.

[0058] (Machine learning methods) 7 is a flowchart showing an example of a machine learning method by the machine learning device 5. In the following, a case will be described in which learning data 13 is used as learning data 13 to generate a learning model 12 as a learning model 12.

[0059] First, in step S100, the learning data acquisition unit 500 acquires a desired number of learning data 13 as a preliminary preparation for starting machine learning, and stores the acquired learning data 13 in the learning data storage unit 52. The number of learning data 13 to be prepared here may be set in consideration of the inference accuracy required for the learning model 12 to be finally obtained.

[0060] Next, in step S110, in order to start machine learning, the machine learning unit 501 prepares a pre-learning learning model 12. The pre-learning learning model 12 prepared here is configured with the neural network model exemplified in Fig. 6, and the weights of each synapse are set to initial values.

[0061] Next, in step S120, the machine learning unit 501 acquires, for example, one set of learning data 13 randomly from the multiple sets of learning data 13 stored in the learning data storage unit 52.

[0062] Next, in step S130, the machine learning unit 501 inputs the simulation specification information (input data) included in one set of learning data 13 to the input layer 120 of the prepared learning model 12 before learning (or during learning). As a result, software setting information (output data) is output as an inference result from the output layer 122 of the learning model 12, and the output data is generated by the learning model 12 before learning (or during learning). Therefore, in the state before learning (or during learning), the output data output as the inference result indicates information different from the software setting information (correct label) included in the learning data 13.

[0063] Next, in step S140, the machine learning unit 501 performs machine learning by comparing the software setting information (correct label) included in the set of learning data 13 acquired in step S120 with the software setting information (output data) output from the output layer as an inference result in step S130 and performing a process (backpropagation) of adjusting the weight of each synapse. In this way, the machine learning unit 501 causes the learning model 12 to learn the correlation between the simulation specification information and the software setting information.

[0064] Next, in step S150, the machine learning unit 501 determines whether a predetermined learning termination condition has been satisfied, for example, based on an evaluation value of an error function based on the software setting information (correct label) included in the learning data 13 and the software setting information (output data) output as an inference result, or the remaining number of unlearned learning data 13 stored in the learning data memory unit 52.

[0065] In step S150, if the machine learning unit 501 determines that the learning end condition is not satisfied and that machine learning is to be continued (No in step S150), the process returns to step S120, and the process of steps S120 to S140 is performed multiple times on the learning model 12 being trained using untrained learning data 13. If the machine learning end condition is satisfied and the machine learning is to be ended (Yes in step S150), the process proceeds to step S160.

[0066] Then, in step S160, the machine learning unit 501 stores the trained learning model 12 (adjusted weight parameter group) generated by adjusting the weights associated with each synapse in the trained model storage unit 53, and ends the series of machine learning methods shown in Fig. 7. In the machine learning method, step S100 corresponds to a learning data storage step, steps S110 to S150 correspond to a machine learning step, and step S160 corresponds to a trained model storage step.

[0067] As described above, the machine learning device 5 and machine learning method of this embodiment can provide a learning model 12 that can predict (infer) software setting information when performing a simulation in accordance with the simulation specification information of the analysis target, from the simulation specification information of the analysis target.

[0068] (Information processing device 6) Fig. 8 is a block diagram showing an example of an information processing device 6 according to an embodiment of the present invention. Fig. 9 is a functional explanatory diagram showing an example of the information processing device 6 according to an embodiment of the present invention. The information processing device 6 includes a control unit 60, a communication unit 61, and a trained model storage unit 62.

[0069] The control unit 60 functions as an information acquisition unit 600, a generation processing unit 601, and an output processing unit 602. The communication unit 61 is connected to an external device via the network 7, and functions as a communication interface for transmitting and receiving various data.

[0070] The information acquisition unit 600 is connected to an external device (e.g., a language processing terminal device 2, an administrator terminal device 3, a software operation support management device 4, etc.) via a communication unit 61 and a network 7, and acquires simulation specification information for the software to be predicted.

[0071] For example, the information acquisition unit 600 refers to the software operation support management database 41 to acquire information related to the space to be analyzed and information related to the medium of the analysis target as simulation specification information of the analysis target in the simulation software to be predicted.

[0072] The generation processing unit 601 generates software setting information for the simulation software based on the software setting information output by inputting the simulation specification information of the analysis target acquired by the information acquisition unit 600 as input data into the learning model 12 as described above.

[0073] The trained model storage unit 62 is a database that stores trained learning models 12 used in the generation processing unit 601. The number of learning models 12 stored in the trained model storage unit 62 is not limited to the above example, and multiple trained models with different conditions may be stored and selectively usable depending on, for example, the machine learning method, the type of target simulation software, the type of data included in the simulation software, the type of data included in the software setting information, etc. Also, the trained model storage unit 62 may be substituted by a storage unit of an external computer (for example, a server-type computer or a cloud-type computer), and in that case, the generation processing unit 601 may access the external computer.

[0074] The output processing unit 602 performs an output process for outputting the software setting information generated by the generation processing unit 601. For example, the output processing unit 602 may transmit the software setting information to the administrator terminal device 3, so that a display screen based on the software setting information is displayed on the administrator terminal device 3, or the software setting information may be used to support software operation. By transmitting the software setting information to the management device 4, the software setting information may be registered in the software operation support management database 41 as a predicted value.

[0075] (Information processing method) 10 is a flowchart showing an example of an information processing method by the information processing device 6. In the following, a case where the learning data 13 is used as the learning model 12 and the learning model 12 is generated as the learning model 12 will be described.

[0076] First, in step S200, the information acquiring unit 600 of the information processing device 6 receives, for example, the type of simulation software to be predicted from the administrator terminal device 3. Then, based on the information from the administrator terminal device 3, the information acquiring unit 600 refers to the software operation support management database 41 to acquire information related to the space to be analyzed and information related to the medium to be analyzed as simulation specification information for the simulation software to be predicted.

[0077] Next, in step S210, the generation processing unit 601 inputs the simulation specification information acquired in step S200 to the learning model 12, thereby generating software setting information for the simulation specification information as output data.

[0078] Next, in step S220, the output processing unit 602 transmits the software setting information generated in step S210 to the administrator terminal device 3 as an output process for outputting the software setting information. The administrator terminal device 3 then displays a display screen based on the software setting information, thereby presenting the software setting information to the administrator. Note that the destination of the software setting information may be the software operation support management device 4 in addition to or instead of the administrator terminal device 3. In the above information processing method, step S200 corresponds to an information acquisition step, step S210 corresponds to a generation processing step, and step S220 corresponds to an output processing step.

[0079] As described above, according to the information processing device 6 and information processing method of the present embodiment, software setting information for simulating the analysis target is predicted by inputting the simulation specification information of the analysis target into the learning model 12. Then, according to such information processing device 6 and information processing method, it becomes possible to execute an optimal simulation in accordance with the simulation specification information of the analysis target.

[0080] (Other embodiments) The present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit and scope of the present invention, all of which are included in the technical concept of the present invention.

[0081] In the above embodiment, the information acquisition unit 600 of the information processing device 6 has been described as acquiring data registered in the software operation support management database 41 as simulation specification information by referring to the software operation support management database 41, but for example, some or all of the data may be acquired directly from the language processing terminal device 2.

[0082] In the above embodiment, the information acquisition unit 600 of the information processing device 6 has been described as acquiring data registered in the software operation support management database 41 as simulation specification information for the analysis target by referring to the software operation support management database 41, but it may also be configured to acquire simulation specification information in which each data (or even a part of it) is provisional data, for example.

[0083] In the above embodiment, the software operation support management device 4, the machine learning device 5, and the information processing device 6 have been described as being configured as separate devices, but these three devices may be configured as a single device, or any two of these three devices may be configured as a single device. Furthermore, at least one of the machine learning device 5 and the information processing device 6 may be incorporated in the language processing terminal device 2 or the administrator terminal device 3.

[0084] In the above embodiment, a case has been described in which a neural network is used as a learning model for implementing machine learning by the machine learning unit 501, but other machine learning models may be used. Examples of other machine learning models include tree types such as decision trees and regression trees, ensemble learning such as bagging and boosting, recurrent neural networks, convolutional neural networks, and neural network types such as LSTM (including deep learning). (including hierarchical clustering, non-hierarchical clustering, k-nearest neighbors, k-means, etc.) Examples of such methods include multivariate analyses such as filtering, principal component analysis, factor analysis, and logistic regression, as well as support vector machines.

[0085] (Machine learning programs and information processing programs) The present invention can also be provided in the form of a program (machine learning program) that causes the computer 900 to function as each unit included in the machine learning device 5, or a program (machine learning program) that causes the computer 900 to execute each step included in the machine learning method. Moreover, the present invention can also be provided in the form of a program (information processing program) that causes the computer 900 to function as each unit included in the information processing device 6, or a program (information processing program) that causes the computer 900 to execute each step included in the information processing method according to the above embodiment.

[0086] (Inference device, inference method, and inference program) The present invention can be provided not only in the form of the information processing device 6 (information processing method or information processing program) according to the above embodiment, but also in the form of an inference device (inference method or inference program) used to infer software setting information when performing a simulation according to simulation specification information of an analysis target. In this case, the inference device (inference method or inference program) can include a memory and a processor, and the processor executes a series of processes. The series of processes includes an information acquisition process (information acquisition step) for acquiring the simulation specification information of the analysis target, and an inference process (inference step) for inferring software setting information when performing a simulation when the simulation specification information is acquired in the information acquisition process.

[0087] By providing it in the form of an inference device (inference method or inference program), it can be easily applied to various devices compared to the case where an information processing device is implemented. It can be naturally understood by those skilled in the art that when the inference device (inference method or inference program) infers software setting information, the inference method implemented by the generation processing unit may be applied using the trained learning model generated by the machine learning device and machine learning method according to the above embodiment.

[0088] The system to which the present invention is applied can also have a function of automatically executing the simulation software and evaluating the calculation process and the calculation results of the analysis software.Specifically, the evaluation items of the calculation process can include, when a calculation is performed with certain settings, whether the analysis starts without any problems (or the settings are wrong and the calculation does not start), whether the analysis starts but the calculation is unstable and diverges and ends before a solution is obtained, whether the analysis starts but ends without obtaining a stable convergent solution, whether the analysis starts and a stable convergent solution is obtained but the time it takes to get to that point is within the user's tolerance range, etc.

[0089] In addition, the calculation results can be evaluated by checking whether they match the correct answer (mathematical exact solution or experimental results, etc.) prepared by the user within the tolerance set by the user. The method of determining whether the answer is correct or not can be any method, such as the degree of agreement with numerical data, various graphs, vector diagrams, streamline diagrams, isosurfaces, and isoline diagrams.

[0090] The system according to the present invention has a function of automatically executing the simulation software, evaluating the calculation process and the calculation results of the simulation software, and further, based on the evaluation results, automatically searching for the optimum simulation software execution method or information that serves as a guideline for determining the execution method (for example, setting A has a short analysis time but medium analysis accuracy, setting B has a long analysis time but high analysis accuracy, etc.), and can also have a function of presenting the results to the user. Reinforcement learning, for example, may be used to search for better settings.

[0091] It is possible to combine the embodiments, and to modify or omit each embodiment as appropriate. [Explanation of symbols]

[0092] 1... software operation support system, 2... language processing terminal device, 3... administrator terminal device, 4...Software operation support management device, 5...Machine learning device, 6...information processing device, 7...network 12...Learning model, 13...Learning data, 17...Natural language processing unit 41...Software operation support management database, 50: control unit, 51: communication unit, 52: learning data storage unit, 53...Trained model memory unit, 60: control unit, 61: communication unit, 62: trained model storage unit, 120...input layer, 121...intermediate layer, 122...output layer 500...learning data acquisition unit, 501...machine learning unit, 600: information acquisition unit; 601: generation processing unit; 602: output processing unit; 900…Computer

Claims

1. an information acquisition unit for acquiring simulation specification information of an analysis target; a generation processing unit that generates the software setting information for the simulation specification information of the analysis target by inputting the simulation specification information acquired by the information acquisition unit into a learning model that has been trained by machine learning on a correlation between the simulation specification information of the analysis target and software setting information when a simulation is performed according to the simulation specification information of the analysis target, Information processing device.

2. The simulation specification information for the analysis target includes: Information about the space being analyzed; and At least one of the pieces of information related to the medium to be analyzed is included. The information processing device according to claim 1 .

3. The software setting information for the simulation includes the following: Model setting information, Condition setting information, and output method setting information, The information processing device according to claim 1 .

4. A natural language processing unit extracts simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software. The information processing device according to claim 1 .

5. An inference device comprising a memory and a processor, The processor, An information acquisition process for acquiring simulation specification information of an analysis target; When the simulation specification information of the analysis target is acquired in the information acquisition process, an inference process is executed to infer software setting information for performing the simulation. Reasoning device.

6. Execute natural language processing to extract simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software.

6. The inference device according to claim 5.

7. a learning data storage unit that stores a plurality of sets of learning data each including simulation specification information of an analysis target and software setting information for performing the simulation; a machine learning unit that inputs a plurality of sets of the learning data into a learning model to cause the learning model to learn a correlation between simulation specification information to be analyzed and software setting information when performing the simulation; A trained model storage unit that stores the trained model in which the correlation is trained by the machine learning unit. Machine learning device.

8. A natural language processing unit extracts simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software. The machine learning device according to claim 7.

9. An information acquisition step of acquiring simulation specification information of an analysis target; a generation processing step of inputting the simulation specification information of the analysis target acquired by the information acquisition step into a learning model that has been trained by machine learning on a correlation between the simulation specification information of the analysis target and software setting information when a simulation is performed according to the simulation specification information of the analysis target, thereby generating software setting information when a simulation is performed. Information processing methods.

10. A natural language processing step is provided for extracting simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software. The information processing method according to claim 9.

11. An inference method executed by an inference device having a memory and a processor, comprising: The processor, An information acquisition process for acquiring simulation specification information of an analysis target; When the simulation specification information of the analysis target is acquired in the information acquisition process, an inference process is executed to infer software setting information for performing the simulation. Reasoning method.

12. Execute natural language processing to extract simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software. The inference method of claim 11.

13. a learning data storage step of storing a plurality of sets of learning data in a learning data storage unit, the learning data being configured with simulation specification information of an analysis target and software setting information for performing the simulation; a machine learning process for inputting a plurality of sets of the learning data into a learning model, thereby causing the learning model to learn a correlation between simulation specification information to be analyzed and software setting information when performing the simulation; and a learned model storage step of storing the learned model in a learned model storage unit, the learned model having learned the correlation by the machine learning step. Machine learning methods.

14. A natural language processing step is provided for extracting simulation specification information to be analyzed and software setting information for performing the simulation from manual document data and / or tutorial document data of the simulation software. The machine learning method of claim 13.

Images