Design support device, design support method, and design support program

Abstract

To provide a design support apparatus, method and program configured to optimize characteristics and design variables of a product or the like, which constitutes objective variables in design process of the product or the like, with a low load.SOLUTION: A design support apparatus 10 includes: a data acquisition unit 11 which acquires actual result data composed of a design parameter group and observation values of characteristic items; a model construction unit 12 which constructs prediction models for predicting observation values of the characteristic items, as a probability distribution or the like, on the basis of the design parameter group; an acquisition function construction unit 13 which constructs a single goal-oriented acquisition function which includes a goal achievement probabilistic term including overall achievement probability that is calculated with the design parameter group as variables on the basis of each of the prediction models and with which target values of all characteristic items are achieved, and receives as input the design parameter group and outputs an index value regarding improvement of characteristics of all characteristic items; a design parameter group acquisition unit 14 which acquires at least one design parameter group by optimizing the goal-oriented acquisition function; and an output unit 15 which outputs design parameter group candidates.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] One aspect of the present disclosure relates to a design support apparatus, a design support method, and a design support program. [Background technology]

[0002] Product design utilizing machine learning has been researched. For example, in the design of functional materials, which is one field of product design, a model is constructed to estimate the properties of a material by machine learning using learning data consisting of pairs of raw material mixture ratios and properties of experimental and already-produced materials, and the properties of the material are predicted for unexperimented raw material mixture ratios. By creating an experiment plan based on such property predictions, it becomes possible to efficiently optimize parameters such as the properties of the material and the raw material mixture ratio, thereby improving development efficiency. In addition, Bayesian optimization is known to be an effective method of such optimization, and a design device that outputs design values ​​using Bayesian optimization is known. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-52737 A Summary of the Invention [Problem to be solved by the invention]

[0004] On the other hand, in the development of products such as materials, in a situation where multiple objective variables (characteristics) are given, multiple objective variables are optimized in order to improve multiple characteristics that change according to design variables. This is called multi-objective optimization. When there is a trade-off between objective variables, there are multiple optimal solutions (Pareto solutions) and it is not determined to be one. For example, when a target value is set for each objective variable, in order to obtain the optimal Pareto solution, it is possible to take an approach of obtaining many Pareto solutions and selecting the Pareto solution that is closest to the design goal. However, such an approach requires the evaluation of many objective functions, which results in a huge processing load and is not realistic. Such problems are not limited to material design but are common to product design in general.

[0005] Therefore, the present invention has been made in consideration of the above problems, and has an object to enable optimization of product characteristics and design variables that constitute target variables in the manufacturing process of a finished product, work-in-progress, semi-finished product, part, or prototype, with less experimentation and less load. [Means for solving the problem]

[0006] A design support device according to one aspect of the present disclosure is a design support device that determines a plurality of design parameters that satisfy a target value set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part, or prototype, in order to be applied to a technique for optimizing design parameters by repeating the process of determining design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on a design parameter group consisting of a plurality of design parameters, the design support device comprising: a data acquisition unit that acquires a plurality of performance data consisting of the design parameter group and each observed value of the plurality of characteristic items for the produced product, work-in-progress, semi-finished product, part, or prototype; The system includes: a model construction unit that constructs a prediction model based on performance data, which predicts an observed value as a probability distribution or its approximation or an alternative index; an acquisition function construction unit that constructs a goal-oriented acquisition function, which is a single acquisition function that takes a group of design parameters as input and outputs index values ​​of the group of design parameters related to improvement of characteristics shown in all characteristic items, the goal-oriented acquisition function including at least a goal achievement probability term including an overall achievement probability, which is the probability that target values ​​of all characteristic items will be achieved and is calculated based on the prediction model using the group of design parameters as variables; a design parameter group acquisition unit that acquires at least one design parameter group by optimizing the goal-oriented acquisition function; and an output unit that outputs the design parameter group acquired by the design parameter group acquisition unit.

[0007] A design support method according to one aspect of the present disclosure is a design support method in a design support device that determines a plurality of design parameters that satisfy a target value set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part, or prototype, in order to be applied to a technique for optimizing design parameters by repeating a process of determining design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on a design parameter group consisting of a plurality of design parameters, the design support method including: a data acquisition step of acquiring a plurality of performance data consisting of a design parameter group and each of the plurality of characteristic items for the produced product, work-in-progress, semi-finished product, part, or prototype; The method includes: a model construction step of constructing a prediction model based on performance data that predicts a value as a probability distribution or an approximation or alternative index thereof; an acquisition function construction unit that constructs a goal-oriented acquisition function that is a single acquisition function that takes a group of design parameters as input and outputs index values ​​of the group of design parameters related to improvement of characteristics shown in all characteristic items, the goal-oriented acquisition function including at least a goal achievement probability term including an overall achievement probability that is the probability that target values ​​of all characteristic items are achieved and is calculated based on the prediction model using the group of design parameters as variables; a design parameter group acquisition step of acquiring at least one design parameter group by optimizing the goal-oriented acquisition function; and an output step of outputting the design parameter group acquired in the design parameter group acquisition step.

[0008] A design support program according to one aspect of the present disclosure is a design support program for causing a computer to function as a design support device that determines a plurality of design parameters that satisfy a target value set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part, or prototype, in order to apply the design parameters to a method of optimizing the design parameters by repeating the process of determining the design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on a design parameter group consisting of a plurality of design parameters, the design support program having the computer function as a design support device that determines a plurality of design parameters that satisfy a target value set for each of a plurality of characteristic items that indicate characteristics of the product, work-in-progress, semi-finished product, part, or prototype, in order to apply the design parameters to a method of optimizing the design parameters by repeating the process of determining the design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on the determined design parameters, the design support program having the computer function of: a data acquisition function that acquires a plurality of performance data consisting of a design parameter group and each observed value of a plurality of characteristic items for the produced product, work-in-progress, semi-finished product, part, or prototype; The system includes a model construction function that constructs, based on performance data, a prediction model that predicts observed values ​​of characteristic items as a probability distribution or an approximation or alternative index thereof based on a group of parameters; an acquisition function construction unit that constructs a goal-oriented acquisition function, which is a single acquisition function that takes a group of design parameters as input and outputs index values ​​of the group of design parameters related to improvement of characteristics shown in all of the characteristic items, wherein the goal-oriented acquisition function includes at least a goal achievement probability term including an overall achievement probability, which is the probability that target values ​​of all of the characteristic items will be achieved and is calculated based on the prediction model using the group of design parameters as variables; a design parameter group acquisition function that acquires at least one design parameter group by optimizing the goal-oriented acquisition function; and an output function that outputs the design parameter group acquired by the design parameter group acquisition function.

[0009] According to this aspect, a prediction model is constructed to predict the observed value of a characteristic item based on performance data. Since this prediction model predicts the observed value as a probability distribution or an approximation or alternative index thereof, it is possible to calculate the achievement probability of the target value of the characteristic item according to a given set of design parameters. In addition, a goal-oriented acquisition function is constructed that receives the set of design parameters as input and outputs an index value related to the fulfillment of the target values ​​of all the characteristic items. Since this goal-oriented acquisition function includes a goal achievement probability term including an overall achievement probability related to the achievement of the target values ​​of all the characteristic items, the overall achievement probability is reflected in the index value output from the goal-oriented acquisition function. Therefore, a set of design parameters capable of achieving the target for the characteristic item can be obtained by optimization using the index value output from the goal-oriented acquisition function as the objective variable.

[0010] In the design support apparatus according to another aspect, the design parameter group acquisition unit may acquire at least one design parameter group that optimizes an output of the goal-oriented acquisition function.

[0011] According to this aspect, it is possible to obtain a group of design parameters that can bring the target regarding the characteristic items closer to being achieved.

[0012] In the design support apparatus according to another aspect, the design parameter group acquisition unit may acquire the plurality of design parameter groups by a predetermined algorithm.

[0013] According to this aspect, a group of multiple design parameters to be used in the next experiment can be easily obtained.

[0014] In a design support device according to another aspect, the overall achievement probability is the sum of the achievement probabilities for the target values ​​of each characteristic item, and the achievement probability for the target value of each characteristic item may be based on a probability distribution of observed values ​​obtained by inputting a group of design parameters into a prediction model for each characteristic item.

[0015] According to this aspect, since the prediction model is configured to output a probability distribution of observed values, it is possible to obtain the achievement probability of the target value of each characteristic item according to the design parameter group. Furthermore, since the overall achievement probability calculated by multiplying the achievement probabilities of the target value of each characteristic item is included in the goal achievement probability term of the goal-oriented acquisition function, the overall achievement probability is appropriately reflected in the index value from the goal-oriented acquisition function.

[0016] In the design support system according to another aspect, the target achievement probability term may be an overall achievement probability or a logarithm of the overall achievement probability.

[0017] According to this aspect, since the goal achievement probability term is made up of the overall achievement probability or the logarithm of the overall achievement probability, the overall achievement probability is appropriately reflected in the index value from the goal-oriented acquisition function.

[0018] In a design support device according to another aspect, the acquisition function construction unit constructs an acquisition function for each characteristic item, the acquisition function taking as input a group of design parameters and outputting an index value of the group of design parameters relating to improvement of the characteristics indicated in the characteristic item, and the goal-oriented acquisition function may further include a term for a weighted sum of the acquisition functions for each characteristic item.

[0019] According to this aspect, a term of the weighted sum of the acquisition functions of each characteristic item is included in the goal-oriented acquisition function, so that improvement in the characteristics of each characteristic item according to the design parameter group is appropriately reflected in the index value from the goal-oriented acquisition function.

[0020] In a design support apparatus according to another aspect, the goal-oriented acquisition function may include a sum of a term of a weighted sum of the acquisition functions of the respective characteristic items and a goal achievement probability term.

[0021] According to this aspect, the degree of improvement in the characteristics for each characteristic item according to the design parameter group and the magnitude of the overall achievement probability regarding the achievement of the target values ​​for all characteristic items are appropriately reflected in the index value from the goal-oriented acquisition function.

[0022] In a design support apparatus according to another aspect, the goal-oriented acquisition function may include a product of a term of a weighted sum of the acquisition functions of the respective characteristic items and a goal achievement probability term.

[0023] According to this aspect, the degree of improvement in the characteristics for each characteristic item according to the design parameter group and the magnitude of the overall achievement probability regarding the achievement of the target values ​​for all characteristic items are appropriately reflected in the index value from the goal-oriented acquisition function.

[0024] In a design support device according to another aspect, the acquisition function construction unit may construct an acquisition function for each characteristic item using any one of a Lower Confidence Bound (LCB), an Expected Improvement (EI), and a Probability of Improvement (PI).

[0025] According to this aspect, an acquisition function suitable for evaluating the improvement of the characteristics shown in each characteristic item is constructed.

[0026] In another aspect of the design support device, the acquisition function construction unit may construct an acquisition function that includes a cost value related to costs including at least one of the time and expenses related to the production of a finished product, a work-in-progress, a semi-finished product, a part, or a prototype, which arise according to a group of design parameters, and outputs an index value indicating that the degree of suitability of the group of design parameters has been reduced as the cost value increases.

[0027] According to this aspect, the cost of manufacturing the product is taken into consideration when acquiring the design parameter group, which makes it possible to reduce the costs of manufacturing the product and conducting experiments.

[0028] In a design support device according to another aspect, the predictive model is a regression model or a classification model that takes a set of design parameters as input and outputs a probability distribution of observed values, and the model construction unit may construct the predictive model by machine learning using performance data.

[0029] According to this aspect, since the prediction model is constructed as a predetermined regression model or classification model, a prediction model is obtained that can obtain a probability distribution of observed values ​​of a characteristic item or an approximation or alternative index thereof.

[0030] In another aspect of the design support device, the prediction model may be a machine learning model that predicts the probability distribution of observed values ​​or an approximation or alternative indicator thereof using any one of a posterior distribution of predicted values ​​based on Bayes' theorem, a distribution of predicted values ​​of predictors constituting an ensemble, a theoretical formula for the prediction interval and confidence interval of a regression model, Monte Carlo dropout, and a distribution of predictions of a plurality of predictors constructed under different conditions.

[0031] According to this aspect, a prediction model is constructed that is capable of predicting the probability distribution of observed values ​​of characteristic items based on a group of design parameters, or an approximation or alternative index thereof. Effect of the Invention

[0032] According to one aspect of the present disclosure, it is possible to optimize the characteristics and design variables of a product or the like that constitute target variables in a manufacturing process of a product, work-in-progress, semi-finished product, part, or prototype with a reduced load and fewer experiments. [Brief description of the drawings]

[0033] [Figure 1] FIG. 1 is a diagram showing an overview of a material design process to which a design support device according to an embodiment is applied. [Diagram 2] 1 is a block diagram showing an example of a functional configuration of a design support apparatus according to an embodiment; [Diagram 3] 1 is a hardware block diagram of a design support apparatus according to an embodiment; [Figure 4] FIG. 13 is a diagram showing an example of a set of design parameters for a fabricated material. [Diagram 5] FIG. 13 shows examples of observed values ​​for fabricated materials. [Figure 6] 1 is a flowchart showing a process of optimizing characteristic items and design parameters in material design. [Figure 7] 4 is a flowchart showing an example of the contents of a design support method in the design support apparatus according to the embodiment. [Figure 8] FIG. 2 is a diagram showing a configuration of a design support program. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0034] Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same or equivalent elements are designated by the same reference numerals, and duplicated description will be omitted.

[0035] FIG. 1 is a diagram showing an overview of a material design process, which is an example of a process for designing a product, work in progress, semi-finished product, part, or prototype to which the design support device according to the embodiment is applied. In the following, "product, work in progress, semi-finished product, part, or prototype" is described as "product, etc." The design support device 10 of the present embodiment can be applied to a process for designing any product, etc. having a plurality of characteristic items indicating the characteristics of the product, etc. and a target value for each characteristic item. The design support device 10 can be applied to a method for optimizing the design parameters and objective variables of a product, etc. by repeating the determination of design parameters and the production of a product, work in progress, semi-finished product, part, or prototype based on the determined design parameters. Specifically, the design support device 10 can be applied to the development and design of materials, as well as the design of products such as automobiles and medicines, and the optimization of the molecular structure of medicines. In this embodiment, as described above, the design support process by the design support device 10 will be described using an example of material design as an example of designing a product, etc.

[0036] 1, the design support process by the design support device 10 is applied to the production and experiment of materials in a plant, a laboratory A, etc. That is, materials are produced in the plant, a laboratory A, etc., based on a set design parameter group x, and observed values ​​y of a plurality of property items indicating the properties of the material are obtained based on the produced material. Note that the material production and experiment in the plant and the laboratory A may be a simulation. In this case, the design support device 10 provides the design parameter group x for the execution of the next simulation.

[0037] The design support device 10 optimizes a plurality of characteristic items and design parameters based on performance data consisting of a design parameter group x and observed values ​​y of a plurality of characteristic items of a material produced based on the design parameter group x. Specifically, the design support device 10 outputs a design parameter group x that may provide more suitable characteristics for the next production and experiment based on the design parameter group x and observed values ​​y for the already produced material.

[0038] For example, the design support device 10 of the present embodiment is applied for the purpose of achieving multiple target properties by tuning multiple design variables in the design of a material product. In one example of designing a material product, when a certain material is produced by mixing multiple polymers and additives, the design support device 10 is used to tune a group of design parameters such as the blending amounts of each polymer and additive as design variables and to tune the group of design parameters so as to achieve target values ​​of multiple property items by using observed values ​​of the property items, such as the elastic modulus and the thermal expansion coefficient, as objective variables.

[0039] Fig. 2 is a block diagram showing an example of a functional configuration of a design support device according to an embodiment. The design support device 10 is a device that obtains a plurality of design parameters that satisfy target values ​​set for a plurality of characteristic items that indicate the characteristics of a material in the design of a material produced based on a design parameter group consisting of a plurality of design parameters. As shown in Fig. 2, the design support device 10 may include a functional unit configured in a processor 101, a design parameter storage unit 21, and an observation value storage unit 22. Each functional unit will be described later.

[0040] 3 is a diagram showing an example of a hardware configuration of a computer 100 constituting the design support device 10 according to the embodiment. The computer 100 can constitute the design support device 10.

[0041] As an example, the computer 100 includes, as hardware components, a processor 101, a main memory device 102, an auxiliary memory device 103, and a communication control device 104. The computer 100 constituting the design support device 10 may further include an input device 105 such as a keyboard, a touch panel, a mouse, or the like, which are input devices, and an output device 106 such as a display.

[0042] The processor 101 is a computing device that executes an operating system and application programs. Examples of processors include a central processing unit (CPU) and a graphics processing unit (GPU), but the type of the processor 101 is not limited to these. For example, the processor 101 may be a combination of a sensor and a dedicated circuit. The dedicated circuit may be a programmable circuit such as a field-programmable gate array (FPGA), or may be another type of circuit.

[0043] The main memory device 102 is a device that stores programs for implementing the design support device 10 and the like, and calculation results output from the processor 101. The main memory device 102 is configured, for example, with at least one of a ROM (Read Only Memory) and a RAM (Random Access Memory).

[0044] The auxiliary storage device 103 is generally a device capable of storing a larger amount of data than the main storage device 102. The auxiliary storage device 103 is configured by a non-volatile storage medium such as a hard disk or a flash memory. The auxiliary storage device 103 stores a design support program P1 for causing the computer 100 to function as the design support device 10, etc., and various data.

[0045] The communication control device 104 is a device that executes data communication with other computers via a communication network, and is configured by, for example, a network card or a wireless communication module.

[0046] Each functional element of the design support device 10 is realized by loading a corresponding program P1 onto the processor 101 or the main memory device 102 and having the processor 101 execute the program. The program P1 includes code for realizing each functional element of the corresponding server. The processor 101 operates the communication control device 104 in accordance with the program P1, and executes reading and writing of data in the main memory device 102 or the auxiliary memory device 103. Through such processing, each functional element of the corresponding server is realized.

[0047] The program P1 may be provided in a form fixedly recorded on a tangible recording medium such as a CD-ROM, a DVD-ROM, a semiconductor memory, etc. Alternatively, at least one of these programs may be provided via a communications network as a data signal superimposed on a carrier wave.

[0048] 2 again, the design support device 10 includes a data acquisition unit 11, a model construction unit 12, an acquisition function construction unit 13, a design parameter group acquisition unit 14, and an output unit 15. The design parameter storage unit 21 and the observation value storage unit 22 may be configured in the design support device 10 as shown in FIG. 2, or may be configured as other devices accessible from the design support device 10.

[0049] The data acquisition unit 11 acquires multiple pieces of performance data regarding already-produced materials. The performance data consists of pairs of design parameters and observed values ​​of multiple characteristic items. The design parameter storage unit 21 is a storage means that stores the design parameters in the performance data, and may be configured in, for example, the main storage unit 102 and the auxiliary storage unit 103. The observed value storage unit 22 is a storage means that stores observed values ​​in the performance data.

[0050] 4 is a diagram showing an example of a design parameter group stored in the design parameter storage unit 21. As shown in FIG. 4, the design parameter storage unit 21 stores a design parameter group x in the first (t=1) to (T-1) times (t=T-1) material fabrication. t The design parameter group x may include, for example, the blending amount of raw material A, the blending amount of raw material B, and a design parameter d, and may constitute vector data with a number of dimensions according to the number of design parameters. In addition to the exemplified design parameters, the design parameters may be non-vector data such as molecular structures and images. In addition, when dealing with a problem of selecting an optimal molecule from a plurality of types of molecules, the design parameters may be data indicating options from among a plurality of molecules.

[0051] 5 is a diagram showing an example of the observed value y stored in the observed value storage unit 22. As shown in FIG. 5, the observed value storage unit 22 stores the observed values ​​y of a plurality of characteristic items (m=1 to M) that indicate the characteristics of the materials produced in the first (t=1) to (T-1) material productions (t=T-1). m,t The characteristic item m may include, for example, a glass transition temperature, an adhesive strength, and a characteristic item M. In addition, each characteristic item has a target value ym(target) The set of design parameters x t and the observed value y m,t The pair constitutes the performance data.

[0052] The design support device 10 calculates whether the observed values ​​of each characteristic item are in line with the target values ​​y based on the performance data from the first (t=1) to (T-1) times (t=T-1) material production. m(target) The observed values ​​of each characteristic item exceed the respective target values ​​y m(target) The Tth design parameter set x T Request.

[0053] The model construction unit 12 constructs a prediction model based on the performance data. The prediction model is a prediction model for determining an observed value y of a characteristic item m based on a set of design parameters x. m A model that predicts the observed value y as a probability distribution or its approximation or alternative indicator. m The model can be any model that can predict the observed value y as a probability distribution or its approximation or alternative index, and the type is not limited. m A prediction model that predicts the probability distribution of observed values ​​using, for example, a distribution of predicted values ​​of predictors that make up an ensemble (random forest), a distribution obtained by Monte Carlo dropout (neural network), or a distribution of predictions of multiple predictors constructed under different conditions (any machine learning method), predicts the probability distribution of observed values ​​as a surrogate indicator.

[0054] For example, a prediction model takes design parameters x as inputs and estimates y m The prediction model may be a regression model that outputs a probability distribution of . When the prediction model is a regression model, the prediction model may be configured by any one of regression models such as Gaussian process regression, random forest, and neural network. The model construction unit 12 may construct the prediction model by a well-known machine learning method using performance data. The model construction unit 12 may construct the prediction model by a machine learning method that applies performance data to the prediction model to update parameters of the prediction model.

[0055] The prediction model may be a machine learning model that predicts the probability distribution of the observed values ​​or its approximation or alternative indicator using any one of the following: the posterior distribution of the predicted values ​​based on Bayes' theorem, the distribution of the predicted values ​​of the predictors constituting the ensemble, the theoretical formula of the prediction interval and confidence interval of the regression model, Monte Carlo dropout, and the distribution of predictions of a plurality of predictors constructed under different conditions. The prediction of the probability distribution of the observed values ​​or its alternative indicator can be obtained by a method specific to the model. The probability distribution of the observed values ​​or its approximation or alternative indicator can be obtained based on the posterior distribution of the predicted values ​​in the case of Gaussian process regression and Bayesian neural network, based on the distribution of predictions of the predictors constituting the ensemble in the case of random forest, based on the prediction interval and confidence interval in the case of linear regression, and based on Monte Carlo dropout in the case of neural network. However, the method of calculating the distribution of the observed values ​​or its alternative indicator for each machine learning model is not limited to the above method.

[0056] In addition, any model can be extended to a model capable of predicting the probability distribution of observed values ​​or its alternative indicators. For example, a model can be exemplified in which a distribution of predicted values ​​of each model obtained by constructing a plurality of data sets by a bootstrap method or the like and constructing a predictive model for each of the data sets is used as an alternative indicator of the probability distribution of observed values. However, the method of extending a machine learning model to a model capable of predicting the probability distribution of observed values ​​or its alternative indicators is not limited to the above method.

[0057] In addition, the predictive model may be constructed using linear regression, PLS regression, Gaussian process regression, bagging ensemble learning such as random forest, boosting ensemble learning such as gradient boosting, support vector machine, neural network, etc.

[0058] In a prediction model constructed as Gaussian process regression, a set of design parameters x in the performance data constituting the explanatory variables of the training data, an observed value y constituting the objective variable, and a design parameter x to be predicted are input into the model to predict the probability distribution of the observed values.

[0059] Furthermore, the model construction unit 12 may tune the hyperparameters of the prediction model by a well-known hyperparameter tuning method. That is, the model construction unit 12 may update the hyperparameters of the prediction model constructed by Gaussian process regression by maximum likelihood estimation using a vector representing a design parameter group x, which is an explanatory variable in the performance data, and an observed value y, which is an objective variable.

[0060] The prediction model may be constructed using a classification model. When the prediction model is a classification model, the model construction unit 12 can construct the prediction model using a machine learning technique that can evaluate a known probability distribution using performance data.

[0061] In this way, the model construction unit 12 constructs a prediction model using a predetermined regression model or classification model, making it possible to acquire a probability distribution of observed values ​​of characteristic items based on an arbitrary set of design parameters x.

[0062] The prediction model may be a single-task model that predicts the observed value of one characteristic item as a probability distribution or its approximation or an alternative index, or a multi-task model that predicts the observed values ​​of multiple characteristic items as a probability distribution or its approximation or an alternative index. In this way, by constructing a prediction model using a multi-task model or single-task model appropriately configured according to the properties of the characteristic items, the accuracy of prediction of the observed value by the prediction model can be improved.

[0063] The acquisition function constructing unit 13 constructs a goal-oriented acquisition function, which is a single acquisition function that receives a group of design parameters as input and outputs index values ​​of the group of design parameters related to improvement of characteristics shown in all characteristic items. The goal-oriented acquisition function includes at least a goal achievement probability term. The goal achievement probability term includes an overall achievement probability, which is a probability calculated based on a prediction model with the group of design parameters as variables. The overall achievement probability is the probability that the target values ​​of all characteristic items are achieved.

[0064] Specifically, the acquisition function constructing unit 13 constructs a goal-oriented acquisition function A'(x) as shown in the following equation (1). A'(x) = g(P(x)) (1) In formula (1), g(P(x)) is a goal achievement probability term. That is, the goal-oriented acquisition function A'(x) includes at least the goal achievement probability term g(P(x)).

[0065] The goal achievement probability term includes the overall achievement probability P(x). For example, the overall achievement probability P(x) may be defined as in the following formula (2) assuming that the goal achievement events of each characteristic item are independent of each other. P(x)=Π 1<=m<=M Pm(x) (2) That is, the overall achievement probability P(x) is the sum of the achievement probabilities Pm(x) of each characteristic item m (m=1 to M). Since the prediction model can predict the probability distribution of the observed values ​​of the characteristic items based on the design parameter set x, the achievement probability Pm(x) of each characteristic item can be expressed as a function with the design parameter set x using the prediction model of each characteristic item as an input variable. Furthermore, the overall achievement probability P(x) may be expressed as a function with the design parameter set x based on the prediction models of all characteristic items as an input variable, without calculating the achievement probability of each characteristic item.

[0066] The goal achievement probability term g(P(x)) is configured to include the overall achievement probability P(x). For example, the goal achievement probability term g(P(x)) may be configured to include the overall achievement probability P(x) as shown in formula (3), or may be configured to include the logarithm of the overall achievement probability P(x) as shown in formula (4). g(P(x))=P(x) (3) g(P(x))=log(P(x)) (4) In addition, the goal achievement probability term may be a term in which the overall achievement probability P(x) or the logarithm of the overall achievement probability P(x) is further multiplied by a coefficient, or may include a term to which other elements are added.

[0067] In this way, in one example of this embodiment, the prediction model is configured to output a probability distribution of observed values, so that it is possible to obtain the achievement probability Pm(x) of the target value of each characteristic item m according to the design parameter group. The overall achievement probability calculated by multiplying the achievement probability Pm(x) of the target value of each characteristic item m is included in the goal achievement probability term of the goal-oriented acquisition function, so that the overall achievement probability is appropriately reflected in the index value from the goal-oriented acquisition function.

[0068] The acquisition function constructing unit 13 may construct a goal-oriented acquisition function that further includes a term of a weighted sum of the acquisition functions Am(x) of each characteristic item m in addition to the goal achievement probability term. The term of the weighted sum of the acquisition functions Am(x) is expressed, for example, as in the following formula (5). Σw m Am(x) ···(5) In equation (5), w m is a weight for each characteristic item m, and may be arbitrarily set in advance. In addition, when Am(x) and P(x) include two types of functions, that is, a function suitable for maximization and a function suitable for minimization, one of the two types may be multiplied by −1 to unify Am(x) and P(x) into a function suitable for maximization or a function suitable for minimization.

[0069] The acquisition function constructing unit 13 may construct a goal-oriented acquisition function A'(x) including a sum of a term of a weighted sum of the acquisition functions of the respective characteristic items and the goal achievement probability term, as in equation (6). A'(x) = Σw m A m ( x ) + g ( P ( x ) ) (6)

[0070] Furthermore, the acquisition function constructing unit 13 may construct a goal-oriented acquisition function A'(x) including a product of a term of a weighted sum of the acquisition functions of the respective characteristic items and the goal achievement probability term, as in equation (6). A'(x) = Σw m Am(x) × g(P(x)) (7)

[0071] In order to construct the goal-oriented acquisition function A'(x) as exemplified by Equation (6) and Equation (7), the acquisition function constructing unit 13 may construct an acquisition function for each characteristic item m, which takes the design parameter group as input variables and outputs an index value of the design parameter group related to the improvement of the characteristic shown in the characteristic item. Specifically, the acquisition function constructing unit 13 constructs an acquisition function Am(x) for each characteristic item m based on the prediction model. The acquisition function Am(x) is a function that takes the design parameter group x as input variables and outputs an index value of the design parameter group related to the improvement of the characteristic shown in each characteristic item m. The acquisition function is a function that outputs an index value indicating the degree of suitability (including closeness to the optimal solution or suitability for searching for the optimal solution) of the design parameter group as an explanatory variable as a solution in order to improve the observed value of the characteristic item predicted by the prediction model.

[0072] The acquisition function constructing unit 13 may construct the acquisition function using a well-known function such as LCB (Lower Confidence Bound).

[0073] LCB is used when minimizing the output of the function, and suitable design parameters can be obtained by minimizing the value of LCB. When constructing an acquisition function by LCB, the acquisition function construction unit 13 defines and constructs an acquisition function Am(x) as shown in the following formula (9). A m ( x ) = m ( x ) - a σ ( x ) (9) The above formula for the acquisition function represents the lower limit of the confidence interval when the observed values ​​predicted by the prediction model are assumed to follow a normal distribution, where m(x) is the mean of the prediction, σ(x) is the variance of the prediction, and a is an arbitrary parameter.

[0074] When the prediction model is constructed as a Gaussian regression process, m(x) and σ(x) are obtained by inputting the set of design parameters x in the performance data constituting the explanatory variables of the teacher data, the observed values ​​y constituting the objective variable, and the set of design parameters x to be predicted into the theoretical formula for the posterior distribution of the Gaussian process regression model.

[0075] Moreover, the acquisition function constructing unit 13 may configure the acquisition function Am(x) using well-known functions such as Expected Improvement (EI) and Probability of Improvement (PI).

[0076] The acquisition function construction unit 13 may construct an acquisition function including a cost function cost(x) that defines the cost (time, expense, etc.) required for manufacturing and experimenting with the material using the design parameter set x for each characteristic item. The acquisition function construction unit 13 constructs an acquisition function that outputs an index value indicating that the suitability of the design parameter set x is reduced as the cost value calculated by the cost function increases.

[0077] Specifically, when constructing an acquisition function that is preferably for maximizing the output, the acquisition function construction unit 13 constructs an acquisition function that outputs a smaller index value as the cost value calculated by the cost function increases. The acquisition function construction unit 13 may construct an acquisition function Am(x)' as shown in the following formula (10). Am(x)'=Am(x)-cost(x) ···(10) In addition, when constructing an acquisition function in which the output is preferably minimized, the acquisition function construction unit 13 constructs an acquisition function that outputs a larger index value as the cost value calculated by the cost function increases. The acquisition function construction unit 13 may construct an acquisition function Am(x)' as shown in the following formula (11). Am(x)'=Am(x)+cost(x) ···(11) Note that the acquisition function including the cost function is not limited to the above example, and may include a term in which the index value is multiplied by or divided by the cost function or cost value.

[0078] By optimizing such a goal-oriented acquisition function, including an acquisition function that takes cost into account, the costs associated with material production are taken into account in the production and experimentation of materials, making it possible to reduce the costs associated with material production, experiments, etc.

[0079] The design parameter group acquisition unit 14 acquires at least one design parameter group by optimizing the goal-oriented acquisition function. Specifically, as an example, the design parameter group acquisition unit 14 may acquire at least one design parameter group that optimizes the output of the goal-oriented acquisition function. Specifically, the design parameter group acquisition unit 14 performs optimization using the index value output from the goal-oriented acquisition function A'(x) constructed by the acquisition function construction unit 13 as the objective variable, and acquires the design parameter group x as the optimal solution.

[0080] Also, as an example, the design parameter group acquisition unit 14 may acquire a plurality of design parameter groups by a predetermined algorithm. Specifically, the design parameter group acquisition unit 14 may acquire a plurality of design parameter groups by applying a batch Bayesian optimization technique to the goal-oriented acquisition function. The batch Bayesian optimization technique may be, for example, a technique such as Local Penalization, but is not limited thereto.

[0081] The output unit 15 outputs the design parameter group acquired by the design parameter group acquisition unit 14. That is, the output unit 15 outputs the design parameter group obtained based on the performance data in the first (t=1) to (T-1) times (t=T-1) material production as the design parameter group x for the Tth material production. T The output is as follows:

[0082] Furthermore, when multiple design parameter groups are acquired by the design parameter group acquisition unit 14, the output unit 15 outputs the acquired design parameter groups as design parameter groups for N times of material production from the (T-1)th time onward. The design parameter groups for multiple times of material production may be used for simultaneous experiments and material production.

[0083] The output form is not limited, but the output unit 15 outputs the candidate design parameter group by, for example, displaying it on a predetermined display device or storing it in a predetermined storage means.

[0084] FIG. 6 is a flowchart showing the process of optimizing characteristic items and design parameters in material design.

[0085] In step S1, a group of design parameters is acquired. The group of design parameters acquired here is for initial material preparation (experiment), and may be a group of design parameters that is arbitrarily set, or may be a group of design parameters that is set based on an experiment or the like that has already been performed.

[0086] In step S2, a material is prepared. In step S3, observed values ​​of the properties of the prepared material are obtained. A pair of a design parameter group as the preparation conditions in step S2 and the observed values ​​of each property obtained in step S3 constitutes performance data.

[0087] In step S4, it is determined whether a predetermined termination condition is satisfied. The predetermined termination condition is a condition for optimizing the design parameter group and the observed value of the characteristic item, and may be set arbitrarily. The termination condition for optimization may be, for example, reaching a predetermined number of times of fabrication (experiment) and acquisition of observed value, reaching of a target value of the observed value, and convergence of optimization. If it is determined that the predetermined termination condition is satisfied, the optimization process is terminated. If it is not determined that the predetermined termination condition is satisfied, the process proceeds to step S5.

[0088] In step S5, the design support process is performed by the design support device 10. The design support process is a process for outputting a group of design parameters for producing the next material. Then, the process returns to step S1 again.

[0089] In the first cycle of the process cycle consisting of steps S1 to S5, if a plurality of pairs of design parameter groups and observed values ​​of characteristic items are obtained as initial data, the processes of steps S1 to S4 are omitted. If no initial data is obtained, a design parameter group obtained by any method such as experimental design and random search is obtained in step S1. In the second and subsequent cycles of the process cycle, the design parameter group output in step S5 is obtained in step S1.

[0090] Fig. 7 is a flowchart showing an example of the contents of the design support method in the design support device 10 according to the embodiment, and shows the process of step S5 in Fig. 6. The design support method is executed by loading a design support program P1 into the processor 101 and executing the program to realize each of the functional units 11 to 15.

[0091] In step S11, the data acquisition unit 11 acquires a plurality of pieces of performance data on the already-produced material. The performance data consists of pairs of a group of design parameters and respective observed values ​​of characteristic items.

[0092] In step S12, the model construction unit 12 constructs a prediction model based on the performance data.

[0093] In step S13, the acquisition function constructing unit 13 constructs a goal-oriented acquisition function that receives the design parameters as input and outputs the index values ​​of the design parameters related to the improvement of the characteristics shown in all characteristic items. The goal-oriented acquisition function includes at least a goal achievement probability term.

[0094] In step S14, the design parameter group acquisition unit 14 acquires a design parameter group by single-objectively optimizing the output of the goal-oriented acquisition function. Specifically, the design parameter group acquisition unit 14 performs optimization using the index value output from the goal-oriented acquisition function A'(x) constructed by the acquisition function construction unit 13 as the objective variable, and acquires the design parameter group x as the optimal solution.

[0095] In step S15, the output unit 15 outputs the design parameter group candidate selected in step S14 as a design parameter group for the next material production (step S1).

[0096] Next, a design support program for causing a computer to function as the design support device 10 of this embodiment will be described. Fig. 8 is a diagram showing the configuration of the design support program.

[0097] The design support program P1 includes a main module m10 that controls the overall design support processing in the design support device 10, a data acquisition module m11, a model construction module m12, an acquisition function construction module m13, a design parameter group acquisition module m14, and an output module m15. The modules m11 to m15 realize the functions of the data acquisition unit 11, the model construction unit 12, the acquisition function construction unit 13, the design parameter group acquisition unit 14, and the output unit 15.

[0098] The design support program P1 may be transmitted via a transmission medium such as a communication line, or may be stored in a recording medium M1 as shown in FIG.

[0099] According to the design support device 10, the design support method, and the design support program P1 of the present embodiment described above, a prediction model is constructed to predict the observed value of a characteristic item based on actual data. Since this prediction model predicts the observed value as a probability distribution or its approximation or alternative index, it is possible to calculate the achievement probability for the target value of the characteristic item according to a given set of design parameters. In addition, a goal-oriented acquisition function is constructed that receives the set of design parameters as input and outputs an index value related to the fulfillment of the target values ​​of all the characteristic items. Since this goal-oriented acquisition function includes a goal achievement probability term including an overall achievement probability related to the achievement of the target values ​​of all the characteristic items, the overall achievement probability is reflected in the index value output from the goal-oriented acquisition function. Therefore, a set of design parameters capable of achieving the target for the characteristic item can be obtained by optimization using the index value output from the goal-oriented acquisition function as the objective variable.

[0100] The present invention has been described in detail above based on the embodiments. However, the present invention is not limited to the above embodiments. The present invention can be modified in various ways without departing from the gist of the present invention. [Explanation of symbols]

[0101] 10...design support device, 11...data acquisition unit, 12...model construction unit, 13...acquisition function construction unit, 14...design parameter group acquisition unit, 15...output unit, 21...design parameter memory unit, 22...observation value memory unit, 100...computer, 101...processor, 102...main memory device, 103...auxiliary memory device, 104...communication control device, 105...input device, 106...output device, M1...recording medium, m10...main module, m11...data acquisition module, m12...model construction module, m13...acquisition function construction module, m14...design parameter group acquisition module, m15...output module, P1...design support program.

Claims

1. A design support device for determining a plurality of design parameters that satisfy target values ​​set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part, or prototype, in order to apply the design parameters to a method of optimizing the design parameters by repeating a process of determining the design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on a group of design parameters consisting of the plurality of design parameters, the plurality of design parameters being determined ... a data acquisition unit that acquires a plurality of pieces of performance data each composed of the design parameter group and each observed value of the plurality of characteristic items with respect to the completed product, the work-in-process, the semi-finished product, the part, or the prototype; a model construction unit that constructs, based on the performance data, a prediction model that predicts an observed value of the characteristic item as a probability distribution or an approximation or an alternative index thereof based on the group of design parameters; an acquisition function constructing unit that constructs a goal-oriented acquisition function, which is a single acquisition function that receives the design parameter group as an input and outputs index values ​​of the design parameter group related to improvement of characteristics shown in all of the characteristic items, the goal-oriented acquisition function including at least a goal achievement probability term including an overall achievement probability, which is a probability that the target values ​​of all characteristic items are achieved and is calculated based on the prediction model with the design parameter group as a variable; a design parameter group acquisition unit that acquires at least one design parameter group by optimizing the goal-oriented acquisition function; an output unit that outputs the design parameter group acquired by the design parameter group acquisition unit, the acquisition function construction unit constructs, for each of the characteristic items, an acquisition function that receives the group of design parameters as an input and outputs an index value of the group of design parameters related to improvement of the characteristic indicated in the characteristic item; The goal-oriented acquisition function further includes a term of a weighted sum of the acquisition functions of each characteristic item. Design support equipment.

2. The design parameter group acquisition unit acquires at least one design parameter group that optimizes an output of the goal-oriented acquisition function. The design support device according to claim 1 .

3. the design parameter group acquisition unit acquires a plurality of the design parameter groups by a predetermined algorithm; The design support device according to claim 1 .

4. the overall achievement probability is the sum of the achievement probabilities of each characteristic item with respect to the target value, the achievement probability of each characteristic item with respect to the target value is based on a probability distribution of the observed values ​​obtained by inputting the design parameters into the prediction model of each characteristic item; The design support device according to any one of claims 1 to 3.

5. The goal achievement probability term is composed of the overall achievement probability or the logarithm of the overall achievement probability. The design support device according to claim 4.

6. The goal-oriented acquisition function includes a sum of a weighted sum of the acquisition functions of the characteristics and the goal achievement probability term. The design support device according to any one of claims 1 to 5.

7. The goal-oriented acquisition function includes a product of a weighted sum of the acquisition functions of each characteristic item and the goal achievement probability term. The design support device according to any one of claims 1 to 5.

8. The acquisition function construction unit constructs the acquisition function for each characteristic item by any one of LCB (Lower Confidence Bound), EI (Expected Improvement), and PI (Probability of Improvement); The design support device according to any one of claims 1 to 7.

9. the acquisition function construction unit constructs the acquisition function, which includes a cost value related to a cost including at least one of the time and expense related to the production of the product, the work in progress, the semi-finished product, the part, or the prototype, which occurs according to the group of design parameters, and outputs the index value indicating that the suitability of the group of design parameters is reduced as the cost value increases. The design support device according to any one of claims 1 to 8.

10. the prediction model is a regression model or a classification model that receives the group of design parameters as an input and outputs a probability distribution of the observed values, The model construction unit constructs the prediction model by machine learning using the performance data. The design support device according to any one of claims 1 to 9.

11. The prediction model is a machine learning model that predicts a probability distribution of an observed value or an approximation or alternative indicator thereof using any one of the following: a posterior distribution of a predicted value based on Bayes' theorem, a distribution of predicted values ​​of predictors constituting an ensemble, a theoretical formula for a prediction interval and a confidence interval of a regression model, Monte Carlo dropout, and a distribution of predictions of a plurality of predictors constructed under different conditions. The design support device according to claim 10.

12. A design support method for a design support device that determines a plurality of design parameters that satisfy target values ​​set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part, or prototype, in order to apply the method to a technique for optimizing design parameters by repeating a process of determining design parameters and producing the product, work-in-progress, semi-finished product, part, or prototype based on a group of design parameters consisting of a plurality of design parameters, the method comprising: a data acquisition step of acquiring a plurality of pieces of performance data each including the design parameter group and each observed value of the plurality of characteristic items with respect to the completed product, the work-in-process, the semi-finished product, the part, or the prototype; a model construction step of constructing, based on the performance data, a prediction model that predicts an observed value of the characteristic item as a probability distribution or an approximation or an alternative index thereof based on the group of design parameters; an acquisition function construction step of constructing a goal-oriented acquisition function which is a single acquisition function that receives the design parameter group as an input and outputs index values ​​of the design parameter group related to improvement of characteristics shown in all of the characteristic items, the goal-oriented acquisition function including at least a goal achievement probability term including an overall achievement probability which is a probability that the target values ​​of all characteristic items are achieved and is calculated based on the prediction model with the design parameter group as a variable; a design parameter set acquisition step of acquiring at least one design parameter set by optimizing the goal-oriented acquisition function; an output step of outputting the design parameter group acquired in the design parameter group acquisition step, In the acquisition function construction step, an acquisition function is constructed for each of the characteristic items, the acquisition function having the design parameter group as an input and an index value of the design parameter group related to improvement of the characteristic indicated in the characteristic item as an output; The goal-oriented acquisition function further includes a term of a weighted sum of the acquisition functions of each characteristic item. Design support method.

13. A design support program for causing a computer to function as a design support device for determining a plurality of design parameters that satisfy target values ​​set for each of a plurality of characteristic items that indicate characteristics of a product, work-in-progress, semi-finished product, part or prototype, in order to apply the program to a method of optimizing design parameters by repeating a process of determining design parameters and producing a product, work-in-progress, semi-finished product, part or prototype based on a group of design parameters consisting of the plurality of design parameters, the plurality of design parameters being determined ... The computer includes: a data acquisition function for acquiring a plurality of pieces of performance data each composed of the design parameter group and each observed value of the plurality of characteristic items with respect to the completed product, the work-in-progress, the semi-finished product, the part, or the prototype; a model construction function that constructs, based on the performance data, a prediction model that predicts an observed value of the characteristic item as a probability distribution or an approximation or an alternative index thereof based on the group of design parameters; an acquisition function construction unit that constructs a goal-oriented acquisition function, which is a single acquisition function that receives the design parameter group as an input and outputs index values ​​of the design parameter group related to improvement of characteristics shown in all of the characteristic items, the goal-oriented acquisition function including at least a goal achievement probability term including an overall achievement probability, which is a probability that the target values ​​of all characteristic items are achieved and is calculated based on the prediction model with the design parameter group as a variable; a design parameter set acquisition function for acquiring at least one design parameter set by optimizing the goal-oriented acquisition function; an output function for outputting the design parameter group acquired by the design parameter group acquisition function; the acquisition function construction function constructs an acquisition function for each of the characteristic items, the acquisition function taking the group of design parameters as input and outputting an index value of the group of design parameters relating to improvement of the characteristic indicated in the characteristic item; The goal-oriented acquisition function further includes a term of a weighted sum of the acquisition functions of each characteristic item. Design support program.

Images