Information processing method, information processing system, and information processing program

Abstract

This information processing method comprises: acquiring first device characteristic information related to a characteristic of a first device having a multilayer structure and device configuration information related to a configuration of a layer of a second device having a multilayer structure to be predicted; generating first design parameter information indicating one or more first design parameters in a configuration of at least one layer of the second device by using the first device characteristic information and the device configuration information; and outputting the first design parameter information.

Description

Information Processing Method, Information Processing System, and Information Processing Program

[0001] The present disclosure relates to a technique for predicting design parameters of a device.

[0002] Patent Document 1 discloses a system for training a device design network. Specifically, a system using a conditional variational autoencoder combined with an adversarial network for randomly generating a device design in which desired device performance is given as a condition has been started.

[0003] In addition, Non-Patent Document 1 discloses a framework for designing a Hamiltonian having desired physical characteristics by using Hamiltonian automatic differentiation, which is one method of an inverse design model.

[0004] Japanese Patent No. 7378647

[0005] Koji Inui, Yukitoshi Motome, “Inverse Hamiltonian design by automatic differentiation”, [online], March 01, 2023, Communications Physics, Nature Publishing Group, <https: / / doi.org / 10.1038 / s42005-023-01132-0>

[0006] However, the conventional techniques described in Patent Document 1 and Non-Patent Document 1 do not acquire device configuration information indicating the configuration of the device and generate device design parameters using this device configuration information. For this reason, the conventional techniques needed further improvement to perform highly versatile device simulations regardless of the configuration of the device.

[0007] The present disclosure has been made to solve such problems, and provides a technique capable of performing highly versatile device simulations regardless of the configuration of the device.

[0008] An information processing method in one aspect of the present disclosure is an information processing method performed by a computer, comprising: acquiring first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generating first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and outputting the first design parameter information.

[0009] According to this disclosure, highly versatile device simulations can be achieved regardless of the device configuration.

[0010] This is a block diagram showing the overall configuration of the information processing device. This is a diagram showing an example of the first display screen that the acquisition unit displays on the display unit. This is a diagram showing an example of the first device characteristic information entered in the device characteristic information input field. This is a flowchart showing an example of the processing performed by the information processing device according to the embodiment. This is a flowchart showing the details of the processing in Figure 4. This is a diagram showing an example of the data structure of the second design parameter information. This is a diagram showing an example of the third device characteristic information. This is a diagram showing another example of the third device characteristic information. This is a diagram showing an example of the second display screen that the output unit displays on the display unit. This is a flowchart showing an example of the processing performed in the information processing device according to embodiment 2. This is a diagram showing an example of the data structure of a plurality of second design parameter information. This is a diagram showing a log of the dialogue that took place between the processing unit and the user using an interactive user interface.

[0011] (Background to one aspect of this disclosure) Device simulators are known that predict device design parameters from device characteristic information that describes the characteristics of the device. The technique of predicting device design parameters from device characteristic information is called inverse design, and the mathematical models used in inverse design are called inverse design models.

[0012] The above device simulator builds its inverse design model based on the assumption that the device configuration is fixed. Therefore, when predicting design parameters for devices with different configurations, the above device simulator requires reconstructing the inverse design model, resulting in low versatility.

[0013] The present inventors have found that by constructing an inverse design model using device configuration information that shows the configuration of a device and predicting the design parameters of the device, highly versatile device simulations can be realized regardless of the device configuration, and have arrived at each aspect of this disclosure.

[0014] (1) An information processing method in one aspect of the present disclosure is an information processing method performed by a computer, which includes: acquiring first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generating first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and outputting the first design parameter information.

[0015] In this configuration, first design parameter information is generated using first device characteristic information and device configuration information regarding the layer configuration of the second device, which is the device to be predicted. Therefore, this configuration enables highly versatile device simulation regardless of the device configuration. In other words, it can predict the design parameters of various devices with different configurations.

[0016] (2) In the information processing method described in (1) above, acquiring the first device characteristic information and the device configuration information may include acquiring design parameter type information relating to the type of the first design parameter together with the first device characteristic information and the device configuration information.

[0017] This configuration makes it possible to generate first design parameter information that indicates a first design parameter of a type corresponding to the type specified in the design parameter type information.

[0018] (3) In the information processing method described in (2) above, acquiring the first device characteristic information may include performing noise reduction processing on the first device characteristic information to acquire second device characteristic information which includes one or more predicted data regarding the characteristics of the first device.

[0019] In this configuration, second device characteristic information is generated by applying noise reduction processing to the first device characteristic information. By using this second device characteristic information, more accurate first design parameter information can be generated.

[0020] (4) In the information processing method described in (2) or (3) above, generating the first design parameter information may include generating second design parameter information that indicates one or more second design parameters that are candidates for the first design parameter, based on the second device characteristic information, the device configuration information, and the design parameter type information.

[0021] According to this configuration, second design parameter information can be generated that indicates one or more second design parameters that are candidates for the first design parameter, based on second device characteristic information, device configuration information, and design parameter type information.

[0022] (5) In the information processing method described in (4) above, generating the first design parameter information includes generating third device characteristic information, which includes predictive data relating to the characteristics of the second device, based on the second design parameter information and the device configuration information, and outputting the first design parameter information may include outputting the third device characteristic information together with the first design parameter information.

[0023] With this configuration, third device characteristic information, which indicates the characteristics of the second device to be predicted, is output along with first design parameter information, allowing the user to understand the characteristics of the second device.

[0024] (6) In the information processing method described in (5) above, the second device characteristic information is data indicating the characteristics of the first device, the third device characteristic information is data indicating the characteristics of the second device, and generating the first design parameter information may include obtaining a difference evaluation value indicating the degree of difference between the data indicated by the second device characteristic information and the data indicated by the third device characteristic information, and determining whether the difference evaluation value is less than or equal to a predetermined threshold.

[0025] This configuration allows for the determination of whether the difference evaluation value, which represents the difference between the data indicated by the second device characteristic information and the data indicated by the third device characteristic information, is below a predetermined threshold. In other words, it allows for the determination of whether the second device characteristic information and the third device characteristic information are similar.

[0026] (7) In the information processing method described in (6) above, generating the first design parameter information may include updating the second design parameter information based on the difference evaluation value when it is determined that the difference evaluation value is greater than the predetermined threshold, and determining the second design parameter information as the first design parameter information when it is determined that the difference evaluation value is less than or equal to the predetermined threshold.

[0027] In this configuration, the second design parameter information used to generate the third device characteristic information, which is determined to have a differential evaluation value below a predetermined threshold, is determined to be the first design parameter information. In other words, if a third device characteristic information similar to the second device characteristic information can be generated, the second design parameter information used to generate that third device characteristic information is determined to be the first design parameter information. This prevents unrealistic second design parameter information from being determined as the first design parameter information.

[0028] (8) In the information processing method described in any one of (5) to (7) above, the second device characteristic information is data indicating the characteristics of the first device, the third device characteristic information is data indicating the characteristics of the second device, and generating the first design parameter information may include generating a plurality of second design parameter information, generating a plurality of third device characteristic information based on each of the plurality of second design parameter information and the device configuration information, obtaining a plurality of difference evaluation values ​​indicating the degree of difference between each of the data indicated by the plurality of third device characteristic information and the data indicated by the second device characteristic information, and determining the second design parameter information corresponding to the third device characteristic information for which the difference evaluation value is smallest as the first design parameter information.

[0029] In this configuration, the second design parameter information corresponding to the third device characteristic information that yields the smallest differential evaluation value is determined as the first design parameter information. In other words, the second design parameter information used to generate the third device characteristic information that is most similar to the second device characteristic information is determined as the first design parameter information. This prevents unrealistic second design parameter information from being determined as the first design parameter information.

[0030] (9) In the information processing method described in any one of (1) to (8) above, the first device characteristic information may include one or more experimental data showing the characteristics of the first device.

[0031] With this configuration, the first design parameter information is generated based on experimental data showing the characteristics of the first device, thus enabling the generation of the first design parameter information more accurately.

[0032] (10) In the information processing method described in any one of (1) to (9) above, the first device and the second device may be energy storage devices.

[0033] This configuration makes it possible to generate first design parameter information that indicates one or more design parameters in the configuration of the energy storage device to be predicted.

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

[0035] (11) An information processing system according to another aspect of the present disclosure is an information processing system including a processor, the processor performing the following: acquiring first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generating first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and outputting the first design parameter information.

[0036] (12) An information processing program according to yet another aspect of the present disclosure causes a computer to acquire first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generate first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and output the first design parameter information.

[0037] This disclosure can also be implemented as an information processing system operated by such an information processing program. Furthermore, it goes without saying that such a program can be distributed via a computer-readable non-temporary recording medium such as a CD-ROM, or via a communication network such as the Internet.

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

[0039] (Embodiment 1) Figure 1 is a block diagram showing the overall configuration of the information processing system 1 in Embodiment 1 of the present disclosure. The information processing system 1 is composed of, for example, a personal computer. However, this is just an example, and the information processing system 1 may be implemented by a server (cloud server, etc.) composed of one or more computers. Alternatively, the information processing system 1 may include a terminal and a server connected to the terminal via a network.

[0040] As shown in Figure 1, the information processing system 1 includes an input unit 10, a processor 20, a storage unit 30, and a display unit 40.

[0041] The input unit 10 includes, for example, a mouse, keyboard, and touch panel.

[0042] The processor 20 includes, for example, a central processing unit (CPU) and / or an image processing unit (GPU).

[0043] The storage unit 30 includes a non-volatile rewritable storage device such as a hard disk drive or a solid-state drive.

[0044] The display unit 40 includes a display such as an LCD monitor.

[0045] The processor 20 includes an acquisition unit 21, a processing unit 22, and an output unit 23. The acquisition unit 21 to the output unit 23 are realized by the processor 20 executing an information processing program stored in the storage unit 30. However, the acquisition unit 21 to the output unit 23 may be constituted by a dedicated hardware circuit such as an ASIC.

[0046] The acquisition unit 21 acquires various information including the first device characteristic information 101 and device configuration information, which will be described later. FIG. 2 is a diagram showing an example of the first display screen 41 (user interface) that the acquisition unit 21 causes to be displayed on the display unit 40. As shown in FIG. 2, the first display screen 41 includes a device characteristic information input field A1, a device configuration information input field A2, a measurement condition information input field A3, and a design parameter type information input field A4.

[0047] The device characteristic information input field A1 accepts the input of first device characteristic information 101 regarding the characteristics of the first device. The first device is a device that serves as a basis for predicting the first design parameter information described later. The first device may be an existing device or a device generated by using a predetermined device simulator capable of virtually generating a device. The first device is a device having a multilayer structure. The first device is, for example, composed of a power storage device. A power storage device is a device that can store electrical energy and can discharge the stored electrical energy as needed. Examples of power storage devices include capacitors such as electric double layer capacitors and solid capacitors, primary batteries, secondary batteries, batteries such as all-solid-state batteries, and the like. However, the configuration of the first device is not limited to power storage devices. The first device may be composed of devices such as solar cells, fuel cells, sensors, thermoelectric conversion devices, light-emitting devices, induction heating devices, spin devices, memory devices, transistors, photodiodes, and the like. Also, the first device may be composed of non-electrical devices such as multilayer films. In addition to these examples, any device having a multilayer structure may correspond to the first device. In the present embodiment, the first device will be described as an existing capacitor.

[0048] The user operates the input unit 10 (such as a mouse, keyboard, etc.) to input the first device characteristic information 101 into the device characteristic information input field A1. For example, the user inputs experimental data indicating the characteristics of the first device into the device characteristic information input field A1.

[0049] Figure 3 shows an example of first device characteristic information 101 entered into the device characteristic information input field A1. More specifically, Figure 3 is a graph showing the change in voltage over time when current is passed through the first device. More specifically, Figure 3 is time-series data showing the change in voltage in the first device when a current of a predetermined magnitude (e.g., 100A) is applied to the first device until a predetermined time has elapsed, and then, at the time the predetermined time has elapsed, a current of the same magnitude (e.g., -100A) is applied in the reverse direction. The first device characteristic information 101 is obtained, for example, by a user of the information processing system 1 conducting an experiment using the first device.

[0050] Furthermore, the first device characteristic information 101 does not necessarily have to be obtained through experimentation. For example, the first device characteristic information 101 may be obtained by having a predetermined device simulator or a machine learning-prepared learning model predict the characteristics of the first device. In addition, the first device characteristic information 101 may be virtual data generated by a user of the information processing system 1 predicting the characteristics of the first device based on their own knowledge.

[0051] The graph shown in Figure 3 is merely one example of the first device characteristic information 101. In addition to the graph shown in Figure 3, various other data can be entered into the device characteristic information input field A1. For example, data showing the I-V characteristics (current-voltage characteristics) of the first device, data showing the charge-discharge characteristics of the first device, data showing the impedance characteristics, data showing the electric field and polarization characteristics, data showing the sensitivity characteristics (ratio of output signal to input signal) of the sensor, data showing the thermoelectric conversion efficiency (relationship with Seebeck coefficient and Joule heating) of the thermoelectric conversion device, data showing the luminescence intensity and wavelength characteristics of the light-emitting device, data showing the heating rate and temperature distribution of the induction heating device, data showing the spin polarization rate of the spin device, data showing the write / read speed and durability of the memory device, data showing the output characteristics (current-voltage characteristics) of the transistor, and data showing the response time and quantum efficiency of the photodiode can all be entered as the first device characteristic information 101. The data format may be a two-dimensional vector format such as spectral data, a one-dimensional vector format, or a scalar format. Furthermore, a user of the information processing system 1 may input one or more pieces of first device characteristic information 101.

[0052] Referring again to Figure 2, the device configuration information input field A2 accepts input of device configuration information regarding the layer configuration of the second device, which is a multilayer device and is the target of prediction by the information processing system 1. Specifically, the user inputs information indicating the components of the second device and information indicating the arrangement of the components of the second device into the device configuration information input field A2. In the example shown in Figure 2, the device configuration information input field A2 contains the information "current collector layer / electrode layer / coat layer / electrolyte layer / electrode layer / current collector layer". That is, information is entered indicating that the second device includes components arranged in the order of current collector layer, electrode layer, coat layer, electrolyte layer, electrode layer, and current collector layer. The symbols used to distinguish layers may be " / ", "-", a space, or a line break. The user can arbitrarily change the content entered in the device configuration information input field A2.

[0053] The measurement condition information input field A3 accepts input of measurement condition information. The measurement condition information indicates the measurement conditions under which the first device characteristic information 101 was acquired. In the example shown in Figure 2, information indicating that the first characteristic information was acquired by setting the current applied to the first device to 100A is entered as measurement condition information. The measurement condition information shown in Figure 2 is just one example, and various other types of information can also be entered into the measurement condition information input field A3.

[0054] The design parameter type information input field A4 accepts input of design parameter type information related to the type of the first design parameter, which will be described later. As will be described in detail later, when the processing unit 22 generates the first design parameter, it generates the first design parameter of the type corresponding to the type specified in the design parameter type information. The types of the first design parameter include, for example, "thickness," "effective ionic conductivity," "effective electronic conductivity," and "porosity." In the design parameter type information input field A4 shown in Figure 2, checkboxes corresponding to thickness, effective ionic conductivity, effective electronic conductivity, etc., are selected. In this case, when the processing unit 22 generates the first design parameter, it generates the thickness of each component included in the second device, the effective ionic conductivity of each component, the effective electronic conductivity of each component, etc., as the first design parameter. In addition, the processing unit 22 generates various types of first design parameters according to the design parameter type information selected (input) by the user.

[0055] The acquisition unit 21 acquires various information entered in the device characteristic information input field A1 to the design parameter type information input field A4. The acquisition unit 21 then inputs this information to the processing unit 22.

[0056] The processing unit 22 generates first design parameter information using at least the first device characteristic information 101 and the device configuration information. The first design parameter information is information indicating one or more first design parameters in the configuration of at least one layer of the second device. For example, consider the case where the second device includes an "electrolyte layer" as a component. In this case, the thickness of the electrolyte layer of the second device, the effective ionic conductivity of the electrolyte layer, and the effective electronic conductivity of the electrolyte layer are the first design parameters. What components the second device includes and what types of parameters are generated as the first design parameters are specified in the device configuration information and the design parameter information.

[0057] The output unit 23 outputs the first design parameter information generated by the processing unit 22. In this embodiment, the output unit 23 displays the first design parameter information on the display unit 40 (display). However, this is just one example. The output unit 23 may also output the first design parameter information to an external device (such as an external server) not shown, which is connected to the information processing system 1 via a communication network such as the Internet. Alternatively, the output unit 23 may output the first design parameter information to the storage unit 30.

[0058] Next, we will explain the processing flow performed by the information processing system 1. Figure 4 is a flowchart showing an example of the processing performed by the information processing system 1 according to Embodiment 1.

[0059] (Step S1) The acquisition unit 21 acquires the first device characteristic information 101 and the device configuration information.

[0060] (Step S2) The processing unit 22 generates first design parameter information based on the first device characteristic information 101 and the device configuration information.

[0061] (Step S3) The output unit 23 generates first design parameter information.

[0062] Figure 5 is a flowchart detailing the process shown in Figure 4. Steps S11 and S12 in Figure 5 correspond to step S1 in Figure 4. Steps S21 to S26 in Figure 5 correspond to step S2 in Figure 4. Step S31 in Figure 5 corresponds to step S3 in Figure 4.

[0063] (Step S11) The acquisition unit 21 generates second device characteristic information from the first device characteristic information 101. Specifically, the acquisition unit 21 displays the first display screen 41 shown in Figure 2 on the display unit 40. The acquisition unit 21 receives input of the first device characteristic information 101 from the user. For example, time-series data as shown in Figure 3 is input as the first device characteristic information 101. The acquisition unit 21 performs a predetermined noise reduction process on the first device characteristic information 101. An example of the noise reduction process is a noise reduction process using a moving average filter. This removes measurement errors contained in the first device characteristic information 101. In this way, the acquisition unit 21 generates second device characteristic information that includes one or more predicted data regarding the characteristics of the first device. In this example, the acquisition unit 21 generates second device characteristic information based on the first characteristic information, which is time-series data. Therefore, the second device characteristic information is time-series data.

[0064] (Step S12) The acquisition unit 21 acquires the second device characteristic information, device configuration information, measurement condition information, and design parameter type information. Specifically, the acquisition unit 21 acquires the information entered in the device configuration information input field A2, the measurement condition information input field A3, and the design parameter type information input field A4 included in the first display screen 41 as device configuration information, measurement condition information, and design parameter type information. The acquisition unit 21 also acquires the second device characteristic information generated in step S11 as the first device characteristic information 101.

[0065] (Step S21) The processing unit 22 generates second design parameter information 200. Specifically, the processing unit 22 generates second design parameter information 200 that indicates one or more second design parameters that are candidates for the first design parameter, based on the second device characteristic information, device configuration information, measurement condition information, and design parameter type information.

[0066] Figure 6 shows an example of the data structure of the second design parameter information 200 generated by the processing unit 22 in step S21. The second design parameter information 200 shown in Figure 6 includes a first column C1 that indicates the type of the second design parameter and a second column C2 that indicates the predicted value of the second design parameter. Specifically, the first column C1 stores information indicating the type of the second design parameter. The information indicating the type of the second design parameter is, for example, "thickness of the electrolyte layer," "effective ionic conductivity of the electrolyte layer," and "effective electronic conductivity of the electrolyte layer." The type stored in the first column C1 corresponds to the type specified in the design parameter type information.

[0067] The second column C2 stores predicted values ​​corresponding to the type stored in the first column C1. Predicted values ​​corresponding to the type include, for example, "electrolyte layer thickness" being "25 μm", "ionic conductivity of the electrolyte layer" being "0.1 S / m", and "electronic conductivity of the electrolyte layer" being "0.05 S / m". In Embodiment 1, these predicted values ​​are randomly generated by the processing unit 22. However, this is just an example. The processing unit 22 may generate second design parameter information 200 using a predetermined machine learning model. For example, the processing unit 22 may obtain predicted values ​​to be stored in the second column C2 by having a pre-trained generative model or a large-scale language model calculate predicted values ​​of the second design parameters from the second device characteristic information, device configuration information, measurement condition information, and design parameter type information.

[0068] Each record constituting the second design parameter information 200 corresponds to one second design parameter. Therefore, Figure 6 shows 10 second design parameters. Note that the types of second design parameters shown in Figure 6 are just examples. There may be 11 or more types of second design parameters, or 9 or fewer types.

[0069] (Step S22) The processing unit 22 generates third device characteristic information 103, which includes predicted data regarding the characteristics of the second device, based on the second design parameter information 200 and the device configuration information. For example, the processing unit 22 generates the third device characteristic information 103 by performing a device simulation that predicts the device characteristics from the device configuration information, measurement condition information, and the second design parameter information 200. In the device simulation, for example, an electrochemical model is used that solves an equation based on Ohm's law under constraints such as the law of conservation of matter. The device simulation differs depending on the type of first device. For example, if the first device is an electric double-layer capacitor, an electrochemical model of an electric double-layer capacitor is used as the device simulation, and if the first device is a solar cell, a semiconductor model is used as the device simulation. For example, the processing unit 22 can generate the third device characteristic information 103 using the electrochemical model of COMSOL Multiphysics developed by COMSOL Corporation.

[0070] Figure 7 shows an example of third device characteristic information 103. More specifically, Figure 7 is time-series data showing the change in voltage in the second device when a current of a predetermined magnitude (e.g., 100A) is applied to the second device until a predetermined time has elapsed, and then, after the predetermined time has elapsed, a current of the same magnitude (e.g., -100A) is applied in the reverse direction. However, the third device characteristic information 103 shown in Figure 7 is just one example. Figure 8 shows another example of third device characteristic information. More specifically, Figure 8 is a charge / discharge curve (calculated data) of a battery (an example of the second device). By changing the content of the measurement condition information entered by the user in the measurement condition information input field A3 (Figure 2), a charge / discharge curve (third device characteristic information 104) as shown in Figure 8 can also be generated. In addition, various other data can be generated as third device characteristic information.

[0071] As another example of the processing in step S22, the processing unit 22 may calculate the third device characteristic information 103 using a machine learning model. More specifically, the processing unit 22 may input device configuration information and second design parameter information 200 into a pre-trained machine learning model, causing the machine learning model to output the third device characteristic information 103. This machine learning model is, for example, a model generated by machine learning a dataset in which the second design parameter information 200 and device configuration information are used as features and the third device characteristic information is used as training data. Alternatively, the processing unit 22 may output the third device characteristic information 103 using a machine learning model that outputs the third device characteristic information 103 from the second design parameter information 200, device configuration information, and measurement condition information. As the machine learning model, for example, a neural network model can be used.

[0072] (Step S23) The processing unit 22 calculates a difference evaluation value that indicates the degree of difference between the second device characteristic information and the third device characteristic information 103. More specifically, the processing unit 22 calculates a difference evaluation value that indicates the degree of difference between the time series data shown by the second device characteristic information and the time series data shown by the third device characteristic information 103. In this embodiment, a large difference evaluation value indicates that the difference between the second device characteristic information and the third device characteristic information 103 is large (they are not similar).

[0073] The processing unit 22 obtains the difference evaluation value by, for example, using a cross-correlation function or a dynamic time stretching method. The method using the cross-correlation function calculates the degree of overlap between the time series data shown by the second device characteristic information and the time series data shown by the third device characteristic information 103 as the difference evaluation value.

[0074] The method utilizing dynamic time stretching is a technique that evaluates the difference between the two by finding corresponding points between the time series data shown by the second device characteristic information and the time series data shown by the third device characteristic information 103, and calculating the distance between these corresponding points.

[0075] In addition, the processing unit 22 can calculate a difference evaluation value between the second device characteristic information and the third device characteristic information 103 using an appropriate evaluation method. Furthermore, although the method for calculating the difference evaluation value when the second device characteristic information and the third device characteristic information are time-series data has been described here, the data structure of the second device characteristic information and the third device characteristic information is not limited to time-series data.

[0076] (Step S24) The processing unit 22 determines whether the difference evaluation value is less than or equal to a predetermined threshold. If the difference evaluation value is less than or equal to the predetermined threshold (YES in step S24), the processing unit 22 proceeds to step S26. On the other hand, if the difference evaluation value is greater than the predetermined threshold (NO in step S24), the processing unit 22 proceeds to step S25.

[0077] (Step S25) The processing unit 22 updates the second design parameter information 200 based on the differential evaluation value. Specifically, when the processing unit 22 executes the process related to step S25, it updates the second design parameter information 200 using the differential evaluation value, second device characteristic information, device configuration information, measurement condition information, and design parameter type information.

[0078] For example, the processing unit 22 calculates the derivative of the function that takes the difference evaluation value into account using automatic differentiation. The "function that takes the difference evaluation value into account" is a function that takes into account the difference between the second device characteristic information and the third device characteristic information 103. Then, the processing unit 22 uses the calculated derivative to update the second design parameter information 200 so that the function that takes the difference evaluation value into account becomes smaller.

[0079] As another example, the processing unit 22 may update the second design parameter information 200 using a Bayesian optimization method. Specifically, the processing unit 22 first uses the second design parameter information 200 as input to construct a machine learning model that predicts the difference between the second device characteristic information and the third device characteristic information 103. Next, the processing unit 22 uses the second design parameter information 200 as input to predict the predicted value of the difference and the predicted variance. Then, the processing unit 22 updates the second design parameter information 200 so that the acquisition function, which takes into account the predicted value and the predicted variance, reaches an extremum.

[0080] In step S25, the processing unit 22 updates the second design parameter information 200 and proceeds to step S22. In step S22, the processing unit 22 generates the third device characteristic information 103 again using the updated second design parameter information 200. In step S23, the processing unit 22 calculates the difference evaluation value between the third device characteristic information 103 and the second device characteristic information again. In step S24, the processing unit 22 determines whether the difference evaluation value is below a predetermined threshold. The processing unit 22 repeats the processes in steps S22 to S25 until it determines YES in step S24.

[0081] (Step S26) If the decision in step S24 is YES and the process proceeds to step S26, the processing unit 22 determines the second design parameter information 200 as the first design parameter information. In other words, the processing unit 22 determines the second design parameter information 200, which has been optimized by the processing from step S21 to step S25, as the first design parameter information.

[0082] (Step S31) In step S31, the output unit 23 outputs the first design parameter information and the third device characteristic information 103. That is, the output unit 23 outputs to the display unit 40 the second design parameter information 200 determined by the processing unit 22 as the first design parameter information, and the third device characteristic information 103 generated using this second design parameter information 200.

[0083] Figure 9 shows an example of a second display screen 42 that the output unit 23 displays on the display unit 40. As shown in Figure 9, the second display screen 42 includes a third device characteristic information display area A11 and a schematic diagram display area A12.

[0084] The third device characteristic information display area A11 displays the third device characteristic information 103 calculated by the processing unit 22. More specifically, the third device characteristic information display area A11 displays the third device characteristic information 103 for which the differential evaluation value has been determined to be below a predetermined threshold (determined as YES in step S24). The third device characteristic information display area A11 may also display the first device characteristic information 101 in addition to the third device characteristic information 103. This allows the user to easily understand the difference between the first device characteristic information 101 and the third device characteristic information 103. Furthermore, the third device characteristic information display area A11 may also display the second device characteristic information.

[0085] The schematic diagram display area A12 displays a schematic diagram corresponding to the device configuration entered in the device configuration information input area A2 shown in Figure 2. The schematic diagram 300 includes multiple regions corresponding to the components that make up the second device. In this example, the schematic diagram 300 includes regions corresponding to the current collector layer, electrode layer, coating layer, electrolyte layer, electrode layer, and current collector layer.

[0086] The user operates the input unit 10 (such as a mouse) to input an operation to select one of the areas included in the schematic diagram 300. Then, the output unit 23 displays the first design parameter display area A13 on the second display screen 42. For example, suppose that in the schematic diagram 300, the user operates the input unit 10 and selects the area of ​​"current collector layer". In this case, as shown in Figure 9, the output unit 23 displays the first design parameter display area A13 on the second display screen 42, which displays the first design parameters related to the current collector layer. In this example, the first design parameter display area A13 displays predicted values ​​of the first design parameters such as the thickness of the current collector layer, the effective ionic conductivity of the current collector layer, and the effective electronic conductivity of the current collector layer.

[0087] The width of each of the multiple regions constituting the schematic diagram 300 has a length corresponding to the predicted value of each component of the second device. For example, suppose that "12 μm" is generated as the first design parameter indicating the "thickness of the current collector layer," "25 μm" is generated as the first design parameter indicating the "thickness of the electrode layer," and "2 μm" is generated as the first design parameter indicating the "thickness of the coating layer." In this case, as shown in schematic diagram display area A12, the output unit 23 sets the width of the "current collector layer" region to be smaller than the width of the "electrode layer" region. Also, the output unit 23 sets the width of the "current collector layer" region to be larger than the width of the "coating layer" region. In this way, the user can intuitively understand the predicted value (thickness in this example) of each component. However, this is just an example, and the width of each region may be set to the same size, for example.

[0088] In step S31, the output unit 23 may further store the first design parameter information and the third device characteristic information 103 in the storage unit 30.

[0089] According to the information processing system 1 described above, first design parameter information is generated using first device characteristic information and device configuration information indicating the configuration of the second device, which is the device to be predicted. Therefore, highly versatile device simulation can be realized regardless of the device configuration.

[0090] Specifically, in the information processing system 1 according to this embodiment, the user can appropriately change the content of the device configuration information entered in the device configuration information input field A2. In other words, the user can appropriately change the configuration of the second device to be predicted. Here, the information processing system 1 generates first design parameter information using the device configuration information and the first device characteristic information 101. Therefore, if the user changes the content of the device configuration information, the processing unit 22 generates first design parameter information using the changed device configuration information and the first device characteristic information 101. In other words, if the user changes the configuration of the second device, the processing unit 22 generates first design parameter information corresponding to that change. As a result, the information processing system 1 can realize a highly versatile device simulation that can handle various devices. Furthermore, the highly versatile device simulation can accelerate the development of various devices. In addition, since the highly versatile device simulation eliminates the need to rebuild the inverse design model which was required for each device configuration, the cost of rebuilding the inverse design model can be reduced.

[0091] (Embodiment 2) The processing unit 22 according to Embodiment 2 is characterized by generating a plurality of second design parameter information 200.

[0092] In Embodiment 2, components similar to those in Embodiment 1 are denoted by the same reference numerals, and their descriptions are omitted. Furthermore, in Embodiment 2, the block diagram is the same as that shown in Figure 1.

[0093] In Embodiment 2, the processing unit 22 generates a plurality of second design parameter information 200 and generates a plurality of third device characteristic information 103 corresponding to each of the plurality of second design parameter information 200. The processing unit 22 calculates a plurality of difference evaluation values ​​between each of the plurality of third device characteristic information 103 and the second device characteristic information. The processing unit 22 calculates the second design parameter information 200 corresponding to the smallest difference evaluation value among the plurality of difference evaluation values ​​as the first design parameter information.

[0094] Figure 10 is a flowchart showing an example of processing in the information processing system 1 according to Embodiment 2. The processing in steps S101 and S102 in Figure 10 corresponds to the processing in step S1 in Figure 4, the processing in steps S201 to S204 in Figure 10 corresponds to the processing in step S2 in Figure 4, and the processing in step S301 in Figure 10 corresponds to the processing in step S3 in Figure 4.

[0095] The processes performed in steps S101 and S102 of Figure 10 are the same as those performed in steps S11 and S12 of Figure 4, so their explanation is omitted.

[0096] (Step S201) The processing unit 22 comprehensively generates a plurality of second design parameter information 200. Specifically, the processing unit 22 generates a plurality of second design parameter information 200 based on second device characteristic information, device configuration information, measurement condition information, and design parameter type information. Figure 11 is a diagram showing an example of the data structure of the plurality of second design parameter information 200 generated by the processing unit 22 in step S201. When generating a plurality of second parameter information, the processing unit 22 randomly generates predicted values ​​to be stored in the second column C2. This generates second design parameter information 200 in which various combinations of predicted values ​​of the second design parameters are combined. Figure 11 shows three second design parameter information 200, but these are just examples. The processing unit 22 may generate three or more second design parameter information 200 so as to cover various combinations of predicted values ​​of the second design parameters.

[0097] As another example of the processing related to step S201, the processing unit 22 may generate a plurality of second design parameter information 200 using a predetermined generative model. Examples of predetermined generative models include generative adversarial networks (GANs), variational autoencoders (VAEs), and diffusion models. In detail, the processing unit 22 may cause a pre-trained generative model to output a plurality of second design parameter information 200 by inputting second device characteristic information, device configuration information, measurement condition information, and design parameter type information into the generative model.

[0098] (Step S202) The processing unit 22 generates a plurality of third device characteristic information 103. Specifically, the processing unit 22 generates a plurality of third device characteristic information 103 based on the plurality of second design parameter information 200 generated in step S201, device configuration information, measurement condition information, and design parameter type information.

[0099] (Step S203) The processing unit 22 calculates a plurality of differential evaluation values ​​based on each of the plurality of third device characteristic information 103 and the second device characteristic information. More specifically, the processing unit 22 calculates a plurality of differential evaluation values ​​that indicate the degree of difference between each of the time series data indicated by the plurality of third device characteristic information 103 and the time series data indicated by the second device characteristic information. The processing unit 22 may calculate the plurality of differential evaluation values ​​by using the cross-correlation function method or the dynamic time stretching method, similar to step S23 of Embodiment 1.

[0100] (Step S204) The processing unit 22 calculates the second design parameter information 200 that minimizes the differential evaluation value as the first design parameter information. More specifically, the processing unit 22 determines the second design parameter information 200 that corresponds to the third device characteristic information 103 that minimizes the differential evaluation value among a plurality of differential evaluation values ​​as the first design parameter information. In other words, the processing unit 22 determines the second design parameter information 200 that was used to generate the third device characteristic information 103 that is most similar to the second device characteristic information as the first design parameter information.

[0101] The process in step S301 is the same as the process in step S31, so its explanation will be omitted.

[0102] According to the information processing system 1 of Embodiment 2, a plurality of second design parameter information 200 is generated, a plurality of third device characteristic information 103 is generated based on these plurality of second design parameter information 200, and a plurality of difference evaluation values ​​indicating the degree of difference between each of the plurality of third device characteristic information 103 and the second device characteristic information are calculated. Then, the second design parameter information 200 corresponding to the third device characteristic information 103 with the smallest difference evaluation value is determined as the first design parameter information. In other words, the second design parameter information 200 used to generate the third device characteristic information 103 that is most similar to the second device characteristic information is determined as the first design parameter information. According to the information processing system 1 of this embodiment, it is possible to suppress the determination of unrealistic second design parameter information 200 as the first design parameter information.

[0103] This disclosure may be modified as follows:

[0104] (1) When the processing unit 22 determines NO in step S24 of Figure 4, it may determine whether the number of times NO was determined in step S24 is equal to or greater than the reference number before starting the process in step S25. The reference number is not particularly limited, but for example it is set to 10 times. If the number of times NO was determined in step S24 is equal to or greater than the reference number (10 times in this case), the processing unit 22 proceeds to step S26. In step S26, the processing unit 22 determines the last generated second design parameter information 200 (latest second design parameter information 200) as the first design parameter information. Alternatively, the processing unit 22 may determine the second design parameter information 200 used to generate the third device characteristic information 103 with the smallest difference evaluation value with the second device characteristic information through 10 update processes as the first design parameter information.

[0105] (2) The processing unit 22 may generate first design parameter information using a pre-trained machine learning model that takes first device characteristic information 101 and device configuration information as inputs and outputs first design parameter information.

[0106] (3) The processing unit 22 may use a pre-trained generative model using a large-scale language model to display an interactive user interface on the display unit 40. Detailed illustrations are omitted, but the user interface includes a text input field. The user inputs prompts to the processing unit 22 by entering text into the text input field. The processing unit 22 generates a response to the user based on the prompts received from the user.

[0107] Figure 12 shows the first log L1, which is a record of the dialogue between the processing unit 22 and the user using an interactive user interface, and the second log L2, which is a continuation of the first log L1. As shown in the first chat CH1 in Figure 12, the user inputs a prompt requesting that the first design parameter information be generated based on the first device characteristic information 101. At this time, the user inputs the first device characteristic information in CSV format. Recognizing the request from the user based on the first chat CH1 and having acquired the first device characteristic information, the processing unit 22 determines that no device configuration information has been input and generates the second chat CH2 as a response to the first chat CH1. In the second chat CH2, the processing unit 22 predicts the device configuration information based on, for example, device configuration information previously input by the user, and asks the user whether it is possible to generate the first design parameter information using this predicted device configuration information. In the third chat CH3, the user inputs a prompt to correct the device configuration information predicted by the processing unit 22. Based on the instructions entered by the user in the first chat CH1, the first device characteristic information, and the device configuration information entered by the user in the third chat CH3, the processing unit 22 generates the third device characteristic information 103. Then, in the fifth chat CH5, the processing unit 22 presents the third device characteristic information 103 to the user. In the sixth chat CH6, the user requests the output of first design parameter information regarding the thickness of the coating layer. That is, the user has entered design parameter type information regarding "coating layer thickness". Recognizing the contents of the sixth chat CH6, the processing unit 22 outputs the first design parameter information regarding "coating layer thickness" in the seventh chat CH7.

[0108] (4) The information processing system 1 may be pre-configured to predict (generate) third device characteristic information 103 under predetermined measurement conditions and to generate first design parameters of a predetermined type. The information processing system 1 according to this modified example receives input of first device characteristic information and device configuration information from the user. The information processing system 1 then generates first design parameter information based on the first device characteristic information obtained from the user, the device configuration information, pre-configured measurement condition information, and pre-configured design parameter type information.

[0109] (5) In Embodiment 1, an example was described in which the acquisition unit 21 acquires first device characteristic information, device configuration information, measurement condition information, and design parameter type information from the first display screen 41. However, this information may be stored in the storage unit 30 in advance. In other words, various types of information such as the first device characteristic information may be acquired by receiving input from the user, or by reading information that is stored in advance in the storage unit 30, etc. In addition, various types of information may be acquired by any other appropriate method.

[0110] An information processing method in one aspect of this disclosure may be one of those shown in (a) to (d) below.

[0111] (a) A computer-based information processing method comprising: acquiring first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generating first design parameter information using the first device characteristic information and the device configuration information; and outputting the first design parameter information, wherein the first design parameter information includes one or more first design parameter values ​​corresponding to each of at least one layer of the second device.

[0112] (b) In the information processing method described in (a) above, acquiring the first device characteristic information and the device configuration information includes acquiring one or more first design parameters together with the first device characteristic information and the device configuration information, and the one or more first design parameters and the one or more first design parameter values ​​may correspond one to one.

[0113] (c) In the information processing method described in (b) above, acquiring the first device characteristic information may include performing noise reduction processing on the first device characteristic information to acquire second device characteristic information which includes one or more predicted data relating to the characteristics of the first device.

[0114] (d) In the information processing method described in (c) above, generating the first design parameter information may include generating one or more second design parameter values ​​that are candidates for the one or more first design parameter values, based on the second device characteristic information, the device configuration information, and the design parameter type information.

[0115] This disclosure is useful for predicting the design parameters of a device.

[0116] 1: Information processing device 10: Input unit 20: Processor 21: Acquisition unit 22: Processing unit 23: Output unit 30: Storage unit 40: Display unit 41: First display screen 42: Second display screen 101: First device characteristic information 103: Third device characteristic information 104: Third device characteristic information 200: Second design parameter information 300: Schematic diagram A1: Device characteristic information input field A11: Third device characteristic information display field A12: Schematic diagram display field A13: First design parameter display area A2: Device configuration information input field A3: Measurement condition information input field A4: Design parameter type information input field C1: First column C2: Second column

Claims

1. An information processing method performed by a computer, comprising: acquiring first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generating first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and outputting the first design parameter information.

2. The information processing method according to claim 1, wherein acquiring the first device characteristic information and the device configuration information includes acquiring design parameter type information relating to the type of the first design parameter together with the first device characteristic information and the device configuration information.

3. The information processing method according to claim 2, wherein acquiring the first device characteristic information includes performing noise reduction processing on the first device characteristic information to acquire second device characteristic information, which includes one or more predicted data relating to the characteristics of the first device, as the first device characteristic information.

4. The information processing method according to claim 3, wherein generating the first design parameter information includes generating second design parameter information indicating one or more second design parameters that are candidates for the first design parameter, based on the second device characteristic information, the device configuration information, and the design parameter type information.

5. The information processing method according to claim 4, wherein generating the first design parameter information includes generating third device characteristic information, which includes predictive data relating to the characteristics of the second device, based on the second design parameter information and the device configuration information, and outputting the first design parameter information includes outputting the third device characteristic information together with the first design parameter information.

6. The information processing method according to claim 5, wherein the second device characteristic information is data indicating the characteristics of the first device, the third device characteristic information is data indicating the characteristics of the second device, and generating the first design parameter information includes obtaining a difference evaluation value indicating the degree of difference between the data indicated by the second device characteristic information and the data indicated by the third device characteristic information, and determining whether the difference evaluation value is less than or equal to a predetermined threshold.

7. The information processing method according to claim 6, wherein generating the first design parameter information includes updating the second design parameter information based on the difference evaluation value when it is determined that the difference evaluation value is greater than the predetermined threshold, and determining the second design parameter information as the first design parameter information when it is determined that the difference evaluation value is less than or equal to the predetermined threshold.

8. The information processing method according to claim 5, wherein the second device characteristic information is data indicating the characteristics of the first device, the third device characteristic information is data indicating the characteristics of the second device, and generating the first design parameter information includes generating a plurality of second design parameter information, generating a plurality of third device characteristic information based on each of the plurality of second design parameter information and the device configuration information, obtaining a plurality of difference evaluation values ​​indicating the degree of difference between each of the data indicated by the plurality of third device characteristic information and the data indicated by the second device characteristic information, and determining the second design parameter information corresponding to the third device characteristic information for which the difference evaluation value is smallest as the first design parameter information.

9. The information processing method according to any one of claims 1 to 8, wherein the first device characteristic information is information including one or more experimental data showing the characteristics of the first device.

10. The information processing method according to any one of claims 1 to 8, wherein the first device and the second device are energy storage devices.

11. An information processing system including a processor, wherein the processor performs the following actions: acquires first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generates first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and outputs the first design parameter information.

12. An information processing program that causes a computer to perform the following actions: acquire first device characteristic information relating to the characteristics of a first multilayer device and device configuration information relating to the layer configuration of a second multilayer device to be predicted; generate first design parameter information indicating one or more first design parameters in the configuration of at least one layer of the second device using the first device characteristic information and the device configuration information; and output the first design parameter information.

Images