WO2022124075A1 - 設計支援装置、設計支援方法及び設計支援プログラム - Google Patents
設計支援装置、設計支援方法及び設計支援プログラム Download PDFInfo
- Publication number
- WO2022124075A1 WO2022124075A1 PCT/JP2021/042983 JP2021042983W WO2022124075A1 WO 2022124075 A1 WO2022124075 A1 WO 2022124075A1 JP 2021042983 W JP2021042983 W JP 2021042983W WO 2022124075 A1 WO2022124075 A1 WO 2022124075A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- parameter group
- design parameter
- design
- probability
- characteristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06313—Resource planning in a project environment
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q90/00—Systems or methods specially adapted for administrative, commercial, financial, managerial or supervisory purposes, not involving significant data processing
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Definitions
- One aspect of this disclosure relates to design support devices, design support methods and design support programs.
- the present invention has been made in view of the above problems, and optimizes the characteristics and design variables of products and the like that constitute objective variables in the manufacturing process of products, work-in-process, semi-finished products, parts or prototypes.
- the purpose is to enable a low load with a smaller number of experiments.
- the design support device is determined to determine the design parameters in the design of products, in-process products, semi-finished products, parts or prototypes manufactured based on a design parameter group consisting of a plurality of design parameters. Products, work-in-process, semi-finished products, parts or prototypes to be applied to the method of optimizing design parameters by repeating the production of products, work-in-process, semi-finished products, parts or prototypes based on the design parameters.
- a design support device that obtains multiple design parameters that satisfy the target values set for each of the multiple characteristic items that indicate characteristics, and is related to manufactured products, in-process products, semi-finished products, parts, or prototypes.
- a data acquisition unit that acquires multiple actual data consisting of the observation values of the design parameter group and multiple characteristic items, and predicts the observation values of the characteristic items as a probability distribution or its approximation or an alternative index based on the design parameter group. Acquisition that inputs the design parameter group and outputs the index value of the design parameter group related to the improvement of the characteristic shown in the characteristic item, based on the model construction unit that builds the prediction model based on the actual data and at least based on the prediction model. Design parameter group candidate generation that generates multiple design parameter group candidates by multi-objective optimization for the acquisition function construction unit that constructs the function for each characteristic item and the design parameter group that uses the output of each of the multiple acquisition functions as the objective variable.
- the overall achievement probability which is the probability that the target values of all characteristic items will be achieved, based on the probability distribution of the observed values obtained by inputting the design parameter group candidate into the prediction model or its approximation or alternative index. It is provided with a selection unit that calculates for each design parameter group candidate and selects at least one design parameter group candidate having the highest overall achievement probability, and an output unit that outputs the selected design parameter group candidate.
- the design support method is determined to determine design parameters in the design of products, in-process products, semi-finished products, parts or prototypes manufactured based on a design parameter group consisting of a plurality of design parameters. Products, work-in-process, semi-finished products, parts or prototypes to be applied to the method of optimizing design parameters by repeating the production of products, work-in-process, semi-finished products, parts or prototypes based on the design parameters.
- a design support method in a design support device that obtains multiple design parameters that satisfy the target values set for each of the multiple characteristic items that indicate the characteristics, such as manufactured products, in-process products, semi-finished products, parts, or
- a data acquisition step for acquiring a plurality of actual data consisting of a design parameter group and observation values of each of a plurality of characteristic items for a prototype, and a probability distribution or an approximation thereof of the observation values of the characteristic items based on the design parameter group.
- the model construction step to build the prediction model to be predicted as an alternative index based on the actual data, and the index value of the design parameter group related to the improvement of the characteristics shown in the characteristic item by inputting the design parameter group at least based on the prediction model.
- the design support program determines design parameters in the design of products, work-in-progress, semi-finished products, parts or prototypes manufactured based on a design parameter group consisting of a plurality of design parameters.
- a design parameter group consisting of a plurality of design parameters.
- a data acquisition function that acquires multiple actual data consisting of design parameter groups and observation values of each of multiple characteristic items for products, work-in-process, semi-finished products, parts, or prototypes, and characteristic items based on the design parameter group.
- a model construction function that builds a prediction model that predicts the observed value of the above as a probability distribution or its approximation or an alternative index based on actual data, and at least the characteristics shown in the characteristic item with the design parameter group as input based on the prediction model.
- All characteristic items are based on the design parameter group candidate generation function that generates multiple design parameter group candidates and the probability distribution of observed values obtained by inputting the design parameter group candidates into the prediction model or its approximation or alternative index.
- the overall achievement probability which is the probability that the target value will be achieved, is calculated for each design parameter group candidate, and the selection function that selects at least one design parameter group candidate with the highest overall achievement probability and the selected design parameter group candidate are output.
- the output function to be realized.
- a prediction model that predicts the observed values of characteristic items based on actual data is constructed. Since this prediction model predicts the observed value as a probability distribution or an approximation thereof or an alternative index, it is possible to calculate the achievement probability for the target value of the characteristic item according to the given design parameter group candidate.
- an acquisition function is constructed for each characteristic item, and the Pareto solution obtained by multi-objective optimization for a design parameter group whose objective variable is the output of a plurality of acquisition functions can be acquired as a design parameter group candidate. Then, the acquired design parameter group candidate is input to the prediction model of each characteristic item to calculate the overall achievement probability, and at least one design parameter group candidate having the highest overall achievement probability is output. Therefore, it is possible to obtain a group of design parameters that may obtain more suitable characteristics, and it is possible to optimize a plurality of characteristics of the manufactured product.
- the prediction model is a regression model or classification model that inputs a design parameter group and outputs a probability distribution of observed values
- the model construction unit uses machine learning using actual data. May be used to build a prediction model.
- the prediction model is constructed as a predetermined regression model or classification model, a prediction model capable of obtaining a probability distribution of observed values of characteristic items or an approximation thereof or an alternative index can be obtained.
- the prediction model is the posterior distribution of the prediction values based on Bayesian theory, the distribution of the prediction values of the predictors constituting the ensemble, the theoretical formulas of the prediction interval and the confidence interval of the regression model, and Monte Carlo Drop. It may be a machine learning model that predicts a probability distribution of observed values or an approximation thereof or an alternative index using any one of out and the prediction distributions of a plurality of predictors constructed under different conditions. ..
- a prediction model that can be predicted as a probability distribution of observed values of characteristic items based on a design parameter group or an approximation thereof or an alternative index is constructed.
- the prediction model is 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 the probability distribution of the observed values of multiple characteristic items. It may be a multitasking model that predicts as an approximation or an alternative index.
- the prediction model can be constructed by a multitasking model or a singletasking model appropriately configured according to the nature of the characteristic item, so that the accuracy of prediction of the observed value by the prediction model can be improved.
- the design parameter group candidate generation unit generates a plurality of design parameter group candidates or generates a plurality of design parameter group candidates by performing one multi-objective optimization by a predetermined first method of multi-objective optimization.
- the multi-objective optimization by the second method of the multi-objective optimization different from the first method may be performed a plurality of times under different conditions to generate a plurality of design parameter group candidates.
- the multi-objective optimization method can be appropriately adopted, it is possible to obtain a plurality of suitable design parameter group candidates.
- the design parameter group candidate generation unit may apply a genetic algorithm to a plurality of acquisition functions to perform multi-objective optimization for the design parameter group.
- optimization of the design parameter group with the output of each of the plurality of acquisition functions as the objective variable is performed with high accuracy.
- the design parameter group candidate generation unit generates a predetermined objective function for treating the multi-objective optimization as a single-objective optimization based on a plurality of acquisition functions, and the objective function. It is also possible to generate a plurality of design parameter group candidates by performing single-objective optimization for a design parameter group having the output of the above as an objective variable a plurality of times under different conditions.
- the multi-objective optimization for a plurality of acquisition functions is realized, so that the design parameter group candidate can be easily obtained. Is possible.
- the selection unit of each characteristic item is based on the probability distribution of the observed value obtained by inputting the design parameter group candidate into the prediction model of each characteristic item or its approximation or an alternative index.
- the achievement probability for the target value may be calculated, and the overall achievement probability may be calculated for each design parameter group candidate based on the achievement probability of each characteristic item.
- the achievement probability for each characteristic item can be calculated with high accuracy. Then, since the overall achievement probability can be obtained by calculating the achievement probability for each characteristic item, the overall achievement probability for each design parameter group candidate can be easily and accurately calculated.
- the selection unit may select a plurality of design parameter group candidates including the design parameter group candidate having the highest overall achievement probability from the plurality of design parameter group candidates by a predetermined algorithm. good.
- the acquisition function construction unit may construct an acquisition function consisting of one of LCB (Lower Confidence Bound), EI (Expected Impression), and PI (Probability of Improvement). good.
- LCB Lower Confidence Bound
- EI Exected Impression
- PI Probability of Improvement
- an acquisition function suitable for evaluating the improvement of the characteristics shown in each characteristic item is constructed.
- the acquisition function construction unit is at least one of the time and cost related to the production of the product, work in process, semi-finished product, part or prototype generated according to the design parameter group. It is also possible to construct an acquisition function that includes a cost value related to the cost including and outputs an index value indicating that the larger the cost value is, the more suitable the design parameter group is reduced.
- the cost related to the production of a product or the like is taken into consideration when generating a design parameter group candidate. Therefore, it is possible to reduce the costs related to the production of products and experiments and experiments.
- the optimization of the characteristics and design variables of the products constituting the objective variables in the manufacturing process of products, work-in-process, semi-finished products, parts or prototypes can be performed with less load by using a smaller number of experiments. Make it possible.
- FIG. 1 is a diagram showing an outline of a material design process which is an example of a product, work-in-process, semi-finished product, part, or prototype design process to which the design support device according to the embodiment is applied.
- products, work in process, semi-finished products, parts or prototypes will be referred to as “products, etc.”
- the design support device 10 of the present embodiment can be applied to a process of designing any product or the like having a plurality of characteristic items indicating the characteristics of the product or the like and a target value of each characteristic item.
- the design support device 10 is a method for optimizing design parameters and objective variables of a product or the like by repeating the determination of design parameters and the production of products, work-in-process, semi-finished products, parts or prototypes based on the determined design parameters. Can be applied to. Specifically, the design support device 10 can be applied not only to the development and design of materials, but also to, for example, the design of products such as automobiles and chemicals, and the optimization of the molecular structure of chemicals. In this embodiment, as described above, the design support process by the design support device 10 will be described by the example of the material design as an example of the design of the product or the like.
- the design support process by the design support device 10 is applied to the production and experiment of materials in the plant, laboratory A and the like. That is, a material is produced in a plant, a laboratory A, or the like by the set design parameter group x, and observation values y of a plurality of characteristic items indicating the characteristics of the material are acquired based on the produced material.
- the material preparation and experiment in the plant and the laboratory A may be a simulation.
- the design support device 10 provides a design parameter group x for executing the next simulation.
- the design support device 10 optimizes a plurality of characteristic items and design parameters based on actual data consisting of observation values y of a plurality of characteristic items of the material produced based on the design parameter group x and the design parameter group x. conduct. Specifically, the design support device 10 may obtain more suitable properties for the next fabrication and experiment based on the design parameter group x and the observed value y for the manufactured material. Output the group x.
- the design support device 10 of the present embodiment is applied for the purpose of tuning a plurality of design variables and achieving a plurality of target characteristics in the design of a material product.
- a design parameter group such as a blending amount of each polymer and the additive as a design variable and has characteristics. It is used for tuning a group of design parameters that achieves the target values of multiple characteristic items, with the observed values of elastic modulus and coefficient of thermal expansion, which are items, as objective variables.
- FIG. 2 is a block diagram showing an example of the functional configuration of the design support device according to the embodiment.
- the design support device 10 has a plurality of target values set for each of the plurality of characteristic items indicating the characteristics of the material in the design of the material manufactured based on the design parameter group consisting of the plurality of design parameters. It is a device that obtains design parameters.
- the design support device 10 may include a functional unit, a design parameter storage unit 21, and an observation value storage unit 22 configured in the processor 101. Each functional part will be described later.
- FIG. 3 is a diagram showing an example of the hardware configuration of the computer 100 constituting the design support device 10 according to the embodiment.
- the computer 100 may configure the design support device 10.
- the computer 100 includes a processor 101, a main storage device 102, an auxiliary storage device 103, and a communication control device 104 as hardware components.
- the computer 100 constituting the design support device 10 may further include an input device 105 such as a keyboard, a touch panel, and a mouse, which are input devices, and an output device 106 such as a display.
- the processor 101 is an arithmetic unit that executes an operating system and an application program. Examples of the processor include a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), but the type of the processor 101 is not limited to these.
- the processor 101 may be a combination of a sensor and a dedicated circuit.
- the dedicated circuit may be a programmable circuit such as FPGA (Field-Programmable Gate Array), or may be another type of circuit.
- the main storage device 102 is a device that stores a program for realizing the design support device 10 and the like, a calculation result output from the processor 101, and the like.
- the main storage device 102 is composed of, for example, at least one of a ROM (Read Only Memory) and a RAM (Random Access Memory).
- the auxiliary storage device 103 is a device capable of storing a larger amount of data than the main storage device 102 in general.
- the auxiliary storage device 103 is composed of a non-volatile storage medium such as a hard disk or a flash memory.
- the auxiliary storage device 103 stores the design support program P1 for making the computer 100 function as the design support device 10 and the like, and various data.
- the communication control device 104 is a device that executes data communication with another computer via a communication network.
- the communication control device 104 is composed of, for example, a network card or a wireless communication module.
- Each functional element of the design support device 10 is realized by loading the corresponding program P1 on the processor 101 or the main storage device 102 and causing the processor 101 to execute the program.
- the program P1 contains a code for realizing each functional element of the corresponding server.
- the processor 101 operates the communication control device 104 according to the program P1 to read and write data in the main storage device 102 or the auxiliary storage device 103. By such processing, each functional element of the corresponding server is realized.
- the program P1 may be provided after being fixedly recorded on a tangible recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory. Alternatively, at least one of these programs may be provided via a communication network as a data signal superimposed on a carrier wave.
- 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 candidate generation unit 14, a selection unit 15, and an output unit 16.
- the design parameter storage unit 21 and the observation value storage unit 22 may be configured in the design support device 10 or as other devices accessible from the design support device 10. ..
- the data acquisition unit 11 acquires a plurality of actual data regarding the manufactured material.
- the actual data consists of a pair of design parameter groups and observation values of each of a plurality of characteristic items.
- the design parameter storage unit 21 is a storage means for storing the design parameter group in the actual data, and may be configured in, for example, a main storage device 102, an auxiliary storage device 103, or the like.
- the observed value storage unit 22 is a storage means for storing the observed values in the actual data.
- FIG. 4 is a diagram showing an example of a design parameter group stored in the design parameter storage unit 21.
- the design parameter group x may include the blending amount of the raw material A, the blending amount of the raw material B, and the design parameter d, and can form vector data having a number of dimensions according to the number of design parameters.
- the design parameters may be, for example, non-vector data such as a molecular structure and an image, in addition to those exemplified. Further, when dealing with the problem of selecting the optimum molecule from the types of a plurality of molecules, the design parameter may be data indicating an option among the plurality of molecules.
- FIG. 5 is a diagram showing an example of the observed value y stored in the observed value storage unit 22.
- the characteristic item m may include, for example, the glass transition temperature, the adhesive force, and the characteristic item M.
- a target value ym (target) is set for each characteristic item.
- the pair of the design parameter group x t and the observed values ym , t constitutes the actual data.
- the design parameter group x T is a parameter group in which the observed value of each characteristic item satisfies each target value ym (target) , or the observed value of each characteristic item is based on each target value ym (target). It is a group of parameters that come close to each other.
- the model building unit 12 builds a prediction model based on actual data.
- the prediction model is a model that predicts the observed value ym of the characteristic item m as a probability distribution or an approximation thereof or an alternative index based on the design parameter group x.
- the model constituting the prediction model may be any model as long as it can be predicted using the observed value ym as a probability distribution or an approximation thereof or an alternative index, and the type is not limited.
- the prediction model that predicts the observed value ym as an alternative index of the probability distribution is, for example, the distribution of the predicted values of the predictors constituting the ensemble (random forest), the distribution obtained by the Monte Carlo dropout (neural network), and under different conditions. Predict the probability distribution of observed values using the distribution of predictions of multiple predictors (arbitrary machine learning method) as an alternative index.
- the prediction model may be a regression model in which the design parameter x is input and the probability distribution of the observed value ym is output.
- the prediction model may be composed of any one of regression models such as Gaussian process regression, random forest and neural network.
- the model building unit 12 may build a prediction model by a well-known machine learning method using actual data.
- the model building unit 12 may build a prediction model by a machine learning method that applies actual data to the prediction model and updates the parameters of the prediction model.
- prediction models under different conditions, such as the posterior distribution of predicted values based on Bayesian theory, the distribution of predicted values of predictors constituting the ensemble, the theoretical formulas of the prediction interval and confidence interval of the regression model, and the Monte Carlo dropout. It may be a machine learning model that predicts a probability distribution of observed values or an approximation thereof or an alternative index using any one of the prediction distributions of the individually constructed predictors. Predictions of the probability distribution of observations or their approximations or alternative indicators can be obtained by model-specific methods.
- the probability distribution of the observed values or its approximation or alternative index is based on the posterior distribution of the predicted values in the case of Gaussian process regression and Basilian neural networks, and in the case of random forests, based on the distribution of the predictions of the predictors that make up the ensemble.
- linear regression it can be obtained based on the prediction interval and confidence interval, and in the case of neural network, it can be obtained based on the Monte Carlo dropout.
- the method of calculating the distribution of observed values for each machine learning model or its alternative index is not limited to the above method.
- any model may be extended to a model that can predict the probability distribution of observed values or its alternative index.
- a model that uses the distribution of predicted values of each model as an alternative index to the probability distribution of observed values which is obtained by constructing a plurality of data sets by the bootstrap method and constructing a prediction model for each.
- an example is given.
- the method of extending the machine learning model to a model capable of predicting the probability distribution of observed values or its alternative index is not limited to the above method.
- the prediction model may be constructed by 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, or the like.
- the design parameter group x in the actual data constituting the explanatory variables of the teacher data, the observation value y constituting the objective variable, and the design parameter x to be predicted are input to the model.
- the probability distribution of the observed values is predicted.
- the model construction unit 12 may tune the hyperparameters of the prediction model by a well-known hyperparameter tuning method. That is, the model building unit 12 is a hyper of the prediction model constructed by Gaussian process regression by maximum likelihood estimation using the vector representing the design parameter group x which is the explanatory variable in the actual data and the observation value y which is the objective variable. You may update the parameters.
- the prediction model may be constructed by a classification model.
- the model construction unit 12 can construct the prediction model by a machine learning method capable of evaluating a well-known probability distribution using actual data.
- the prediction model is 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 predicts the observed value of a plurality of characteristic items as a probability distribution or its approximation or an alternative index. It may be a multitasking model. In this way, by constructing a prediction model using a multitasking model or a singletasking model appropriately configured according to the nature of the characteristic item, the accuracy of prediction of the observed value by the prediction model can be improved.
- the acquisition function construction unit 13 constructs the acquisition function Am (x) for each characteristic item m based on the prediction model constructed for each characteristic item m.
- the acquisition function Am (x) is a function that inputs the design parameter group x and outputs the index value of the design parameter group related to the improvement of the characteristic shown in each characteristic item m.
- the acquisition function is suitable for improving the observed values of the characteristic items predicted by the prediction model as a solution of the design parameter group (close to the optimum solution or searching for the optimum solution). It is a function that outputs an index value indicating (including being suitable).
- the acquisition function construction unit 13 may construct an acquisition function by a well-known function such as LCB (Lower Connection Bound).
- the LCB is used to minimize the output of the function, and by minimizing the value of the LCB, suitable design parameters can be obtained.
- the acquisition function construction unit 13 defines and constructs the acquisition function Am (x) as follows.
- Am (x) m (x) -a ⁇ (x)
- the equation of the acquisition function is an equation representing the lower limit of the confidence interval when it is assumed that the observed values predicted by the prediction model follow a normal distribution, and m (x) in the above equation is the average of predictions, ⁇ (x). Is the variance of the prediction, and a is an arbitrary parameter.
- the design parameter group x in the actual data constituting the explanatory variables of the teacher data and the observed values constituting the objective variable are added to the theoretical formula of the posterior distribution of the Gaussian process regression model.
- y and the design parameter group x to be predicted By inputting y and the design parameter group x to be predicted, m (x) and ⁇ (x) can be obtained.
- the acquisition function construction unit 13 may configure the acquisition function Am (x) by well-known functions such as EI (Expected Impression) and PI (Probability of Improvement).
- the acquisition function construction unit 13 may construct an acquisition function including the cost function cost (x) that defines the cost (time, cost, etc.) required for the production and experiment of the material by the design parameter group x.
- the acquisition function construction unit 13 constructs an acquisition function that outputs an index value indicating that the appropriate degree of the design parameter group x is reduced as the cost value calculated by the cost function is larger.
- the acquisition function construction unit 13 when constructing an acquisition function in which it is preferable to maximize the output, the acquisition function construction unit 13 outputs a smaller index value as the cost value calculated by the cost function increases.
- Am (x)' Am (x) + cost (x)
- the acquisition function including the cost function is not limited to the above example, and may include a term for multiplying the cost function or the cost value by the index value or dividing the index value by the cost function or the cost value.
- the cost related to the production of the material is taken into consideration in the production and the experiment of the material. Therefore, the cost related to the production and the experiment of the material is considered. Reduction is possible.
- the design parameter group candidate generation unit 14 generates a plurality of design parameter group candidates by multi-objective optimization for the design parameter group whose objective variable is the output of each of the plurality of acquisition functions.
- the design parameter group candidate generation unit 14 may perform multi-objective optimization using any of the well-known methods, and the method is not limited.
- the design parameter group candidate generation unit 14 may generate a plurality of design parameter group candidates by performing multi-objective optimization by a predetermined method a plurality of times under different conditions. Specifically, the design parameter group candidate generation unit 14 generates M acquisition functions (A 1 (x), A 2 (x), ..., Am (x)) constructed for each characteristic item. , Multi-objective optimization is performed on the design parameter group x, and design parameter group candidates are obtained. The design parameter group candidate generation unit 14 obtains a plurality of design parameter group candidates by performing the multi-objective optimization a plurality of times by changing a predetermined parameter in the multi-objective optimization.
- the design parameter group candidate generation unit 14 may generate a plurality of design parameter group candidates by performing one multi-objective optimization by a predetermined method.
- the design parameter group candidate generation unit 14 applies a genetic algorithm to a plurality of acquisition functions to perform multi-objective optimization for the design parameter group.
- a plurality of design parameter group candidates can be obtained by performing one optimization process.
- the genetic algorithm is an algorithm that imitates the evolution of living organisms and optimizes a function by repeating evaluation, selection, and genetic manipulation for an individual (solution candidate) set. Since a plurality of individuals are evolved at the same time in a genetic algorithm, a Pareto solution can be obtained by evolving to approach the Pareto solution while maintaining the diversity of a plurality of solution candidates. There are various genetic algorithms applicable to multi-objective optimization, but any algorithm may be applied.
- the design parameter group candidate generation unit 14 may acquire a plurality of design parameter group candidates by performing optimization by a genetic algorithm based on a randomly generated initial population a plurality of times.
- the design parameter group candidate generation unit 14 in order to generate the design parameter group candidate, the design parameter group candidate generation unit 14 generates a predetermined objective function for treating the multi-objective optimization as a single-objective optimization based on a plurality of acquisition functions. You may.
- the design parameter group candidate generation unit 14 performs single-objective optimization for the design parameter group whose objective variable is the output of the generated objective function a plurality of times by changing the single-objective function generation conditions. Multiple can be generated.
- Arbitrary parameter group w m that satisfies the condition is randomly selected.
- the following objective function g (x) is generated.
- the objective function g (x) is a function that makes a multi-objective problem that maximizes a plurality of acquisition functions into a uni-objective problem that minimizes the objective function g (x).
- the selection unit 15 is the overall achievement probability, which is the probability that the target values of all the characteristic items are achieved based on the probability distribution of the observed values obtained by inputting the design parameter group candidate into the prediction model or its approximation or an alternative index. Is calculated for each design parameter group candidate.
- the selection unit 15 may calculate the overall achievement probability based on the design parameter group candidates by using the prediction model of all the characteristic items. Further, the selection unit 15 inputs the design parameter group candidate into each of the prediction models of each characteristic item, calculates the achievement probability for the target value of each characteristic item, and achieves the whole based on the achievement probability of each characteristic item. You may calculate the probability.
- the prediction model is a model that predicts the observed value ym of the characteristic item m as a probability distribution or an approximation thereof or an alternative index based on the design parameter group x. Therefore, the selection unit 15 inputs the design parameter group candidate generated by the design parameter group candidate generation unit 14 into the prediction model of each characteristic item, so that the probability distribution of the observed value of each characteristic item or its approximation or an alternative index thereof. To get. Then, the selection unit 15 calculates the achievement probability Pm (x) for each characteristic item m based on the probability distribution of the observed value of each characteristic item or its approximation or an alternative index, and the target value.
- the achievement probability Pm (x) is the probability that the observed value y m for the characteristic item m of the material produced by the design parameter group x achieves the target value y m (target) .
- the selection unit 15 calculates the overall achievement probability P (x) as follows.
- the selection unit 15 calculates the overall achievement probability P (x) of each of the plurality of design parameter group candidates x generated by the design parameter group candidate generation unit 14. Then, the selection unit 15 selects at least one design parameter group candidate having the highest calculated overall achievement probability P (x).
- the selection unit 15 may select a plurality of design parameter group candidates including the design parameter group candidate having the highest overall achievement probability from the plurality of design parameter group candidates by a predetermined algorithm.
- the selection unit 15 selects the following first to Nth design parameter group candidates.
- First design parameter group candidate Design parameter group candidate with the highest overall achievement probability among multiple design parameter group candidates
- Second design parameter group candidate The overall achievement probability of the design parameter group candidates excluding the first design parameter group candidate and the design parameter group candidates distributed within a predetermined distance in the vicinity thereof from the plurality of design parameter group candidates is Highest design parameter group candidate
- Third design parameter group candidate Overall achievement of the design parameter group candidates excluding the first and second design parameter group candidates and the design parameter group candidates distributed within a predetermined distance in the vicinity thereof from the plurality of design parameter group candidates.
- Nth design parameter group candidate From a plurality of design parameter group candidates, the first, second, and so on. .. .. The design parameter group candidate having the highest overall achievement probability among the design parameter group candidates excluding the design parameter group candidates of N-1 and the design parameter group candidates distributed within a predetermined distance in the vicinity thereof.
- the method of performing Bayesian optimization by selecting the first to Nth design parameter group candidates as described above is called batch Bayesian optimization.
- the output unit 16 is a design parameter group for producing materials for N times after the (T-1) th time.
- the selected design parameter group candidate is output.
- the design parameter group for N batches of material fabrication may be used for simultaneous experiments and material fabrication.
- the output unit 16 outputs a design parameter group candidate by displaying it on a predetermined display device or storing it in a predetermined storage means, for example.
- FIG. 6 is a flowchart showing the process of optimizing characteristic items and design parameter groups in material design.
- step S1 the design parameter group is acquired.
- the design parameter group acquired here is for initial material fabrication (experiment), may be an arbitrarily set design parameter group, or is set based on an experiment or the like that has already been performed. It may be a set of design parameters.
- step S3 Material production is performed in step S2.
- step S3 the observed value of the characteristic item of the produced material is acquired.
- the pair of the design parameter group as the production condition in step S2 and the observed value of each characteristic item acquired in step S3 constitutes the actual data.
- step S4 it is determined whether or not the predetermined end condition is satisfied.
- the predetermined end condition is a condition for optimizing the observed values of the design parameter group and the characteristic item, and may be arbitrarily set.
- the end condition for optimization may be, for example, the arrival of a predetermined number of preparations (experiments) and acquisition of observed values, the arrival of an observed value at a target value, and the convergence of optimization. If it is determined that the predetermined termination conditions are satisfied, the optimization process is terminated. If it is not determined that the predetermined termination condition is satisfied, the process proceeds to step S5.
- step S5 the design support process by the design support device 10 is performed.
- the design support process is a process of outputting a group of design parameters for manufacturing the next material. Then, the process returns to step S1 again.
- step S5 the design parameter group output in step S5 is acquired in step S1.
- 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.
- the design support method is executed by loading the design support program P1 into the processor 101 and executing the program to realize the functional units 11 to 16.
- step S11 the data acquisition unit 11 acquires a plurality of actual data regarding the manufactured material.
- Actual data consists of pairs of design parameter groups and observed values of characteristic items.
- step S12 the model building unit 12 builds a prediction model based on the actual data.
- step S13 the acquisition function construction unit 13 constructs the acquisition function Am (x) for each characteristic item m based on the constructed prediction model.
- step S14 the design parameter group candidate generation unit 14 performs multi-objective optimization for the design parameter group whose objective variable is the output of each of the plurality of acquisition functions, and acquires a plurality of design parameter group candidates.
- step S15 the selection unit 15 sets the target values of all the characteristic items based on the probability distribution of the observed values obtained by inputting the design parameter group candidate into the prediction model of each characteristic item or its approximation or an alternative index.
- the overall achievement probability which is the probability to be achieved, is calculated for each design parameter group candidate.
- step S16 the selection unit 15 selects at least one design parameter group candidate having the highest overall achievement probability calculated in step S15.
- step S17 the output unit 16 outputs the design parameter group candidate selected in step S16 as a design parameter group for the next material fabrication (step S1).
- FIG. 8 is a diagram showing the structure of the design support program.
- the design support program P1 includes a main module m10 that comprehensively controls 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 candidate generation module m14, and a selection module. It is configured to include m15 and an output module m16. Then, each module m11 to m16 realizes each function for the data acquisition unit 11, the model construction unit 12, the acquisition function construction unit 13, the design parameter group candidate generation unit 14, the selection unit 15, and the output unit 16.
- the design support program P1 may be transmitted via a transmission medium such as a communication line, or may be stored in the recording medium M1 as shown in FIG.
- a prediction model for predicting the observed values of the characteristic items is constructed based on the actual data. Since this prediction model predicts the observed value as a probability distribution or an approximation thereof or an alternative index, it is possible to calculate the achievement probability for the target value of the characteristic item according to the given design parameter group candidate.
- an acquisition function is constructed for each characteristic item, and the Pareto solution obtained by multi-objective optimization for a design parameter group whose objective variable is the output of a plurality of acquisition functions can be acquired as a design parameter group candidate.
- the acquired design parameter group candidate is input to the prediction model of each characteristic item to calculate the overall achievement probability, and at least one design parameter group candidate having the highest overall achievement probability is output. Therefore, it is possible to obtain a group of design parameters that may obtain more suitable characteristics, and it is possible to optimize a plurality of characteristics of the manufactured product.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Geometry (AREA)
- Strategic Management (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Manufacturing & Machinery (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
Am(x)=m(x)-aσ(x)
上記獲得関数の式は、予測モデルにより予測される観測値が正規分布に従うと仮定した場合の信頼区間下限を表す式であって、上記式におけるm(x)は予測の平均、σ(x)は予測の分散、aは任意のパラメータである。
Am(x)’=Am(x)-cost(x)
また、出力が最小化されることが好適な獲得関数を構築する場合には、獲得関数構築部13は、コスト関数により算出されるコスト値が大きいほど、大きい指標値を出力するような獲得関数を構築する。例えば、以下のような獲得関数Am(x)’を構築してもよい。
Am(x)’=Am(x)+cost(x)
なお、コスト関数を含む獲得関数は、上記の例に限られず、コスト関数またはコスト値を指標値に乗じたり、指標値をコスト関数またはコスト値により除したりする項を含んでもよい。
Am’(x)=-Am(x)
g(x)=maxm[wmAm’(x)]+ρΣmwmAm’(x)
目的関数g(x)は、複数の獲得関数を最大化する多目的問題を、目的関数g(x)を最小化するという単目的問題にする関数である。
P(x)=Π1<=m<=MPm(x)
即ち、全体達成確率P(x)は、M個の達成確率を総乗することにより算出される。
第1の設計パラメータ群候補:複数の設計パラメータ群候補のうちの、全体達成確率が最も高い設計パラメータ群候補、
第2の設計パラメータ群候補:複数の設計パラメータ群候補から、第1の設計パラメータ群候補及びその近傍所定距離内に分布する設計パラメータ群候補を除いた設計パラメータ群候補のうちの全体達成確率が最も高い設計パラメータ群候補、
第3の設計パラメータ群候補:複数の設計パラメータ群候補から、第1,2の設計パラメータ群候補及びその近傍所定距離内に分布する設計パラメータ群候補を除いた設計パラメータ群候補のうちの全体達成確率が最も高い設計パラメータ群候補、
・・・、
第Nの設計パラメータ群候補:複数の設計パラメータ群候補から、第1,2,...N-1の設計パラメータ群候補及びその近傍所定距離内に分布する設計パラメータ群候補を除いた設計パラメータ群候補のうちの全体達成確率が最も高い設計パラメータ群候補。
Claims (13)
- 複数の設計パラメータからなる設計パラメータ群に基づいて作製される製品、仕掛品、半製品、部品又は試作品の設計において、設計パラメータの決定と決定された設計パラメータに基づく製品、仕掛品、半製品、部品又は試作品の作製との繰り返しにより設計パラメータの最適化を図る手法に適用するために、製品、仕掛品、半製品、部品又は試作品の特性を示す複数の特性項目のそれぞれについて設定された目標値を満たすような、前記複数の設計パラメータを求める設計支援装置であって、
作製済みの前記製品、前記仕掛品、前記半製品、前記部品又は前記試作品に関しての、前記設計パラメータ群と前記複数の特性項目のそれぞれの観測値とからなる実績データを複数取得するデータ取得部と、
前記設計パラメータ群に基づいて前記特性項目の観測値を確率分布若しくはその近似又は代替指標として予測する予測モデルを、前記実績データに基づいて構築するモデル構築部と、
少なくとも前記予測モデルに基づいて、前記設計パラメータ群を入力とし前記特性項目に示される特性の向上に関する前記設計パラメータ群の指標値を出力とする獲得関数を前記特性項目ごとに構築する獲得関数構築部と、
前記複数の獲得関数のそれぞれの出力を目的変数とする前記設計パラメータ群についての多目的最適化により、設計パラメータ群候補を複数生成する設計パラメータ群候補生成部と、
前記設計パラメータ群候補を前記予測モデルに入力することにより得られる前記観測値の確率分布若しくはその近似又は代替指標に基づいて、全ての特性項目の前記目標値が達成される確率である全体達成確率を前記設計パラメータ群候補ごとに算出し、前記全体達成確率が最も高い少なくとも一つの前記設計パラメータ群候補を選択する選択部と、
選択した前記設計パラメータ群候補を出力する出力部と、
を備える設計支援装置。 - 前記予測モデルは、前記設計パラメータ群を入力とし、前記観測値の確率分布を出力とする回帰モデルまたは分類モデルであり、
前記モデル構築部は、前記実績データを用いた機械学習により、前記予測モデルを構築する、
請求項1に記載の設計支援装置。 - 前記予測モデルは、ベイズ理論に基づく予測値の事後分布、アンサンブルを構成する予測器の予測値の分布、回帰モデルの予測区間及び信頼区間の理論式、モンテカルロドロップアウト、及び、異なる条件で複数個構築した予測器の予測の分布のうちのいずれか一つを用いて観測値の確率分布若しくはその近似又は代替指標を予測する機械学習モデルである、
請求項2に記載の設計支援装置。 - 前記予測モデルは、一の特性項目の観測値を確率分布若しくはその近似又は代替指標として予測するシングルタスクモデル、または、複数の特性項目の観測値を確率分布若しくはその近似又は代替指標として予測するマルチタスクモデルである、
請求項1~3のいずれか一項に記載の設計支援装置。 - 前記設計パラメータ群候補生成部は、
多目的最適化の所定の第1の手法による1回の多目的最適化の実施により、複数の設計パラメータ群候補を生成し、または、
前記第1の手法とは異なる多目的最適化の第2の手法による多目的最適化を、条件を変えて複数回行うことにより、複数の設計パラメータ群候補を生成する、
請求項1~4のいずれか一項に記載の設計支援装置。 - 前記設計パラメータ群候補生成部は、前記複数の獲得関数に遺伝的アルゴリズムを適用して前記設計パラメータ群についての前記多目的最適化を実施する、
請求項5に記載の設計支援装置。 - 前記設計パラメータ群候補生成部は、前記複数の獲得関数に基づいて、前記多目的最適化を単目的最適化として扱うための所定の一の目的関数を生成し、前記目的関数の出力を目的変数とする前記設計パラメータ群についての単目的最適化を、条件を変えて複数回行うことにより、設計パラメータ群候補を複数生成する、
請求項5に記載の設計支援装置。 - 前記選択部は、前記設計パラメータ群候補を各特性項目の前記予測モデルに入力することにより得られる前記観測値の確率分布若しくはその近似又は代替指標に基づいて各特性項目の前記目標値に対する達成確率を算出し、各特性項目の前記達成確率に基づいて、前記全体達成確率を前記設計パラメータ群候補ごとに算出する、
請求項1~7のいずれか一項に記載の設計支援装置。 - 前記選択部は、複数の前記設計パラメータ群候補から、前記全体達成確率が最も高い前記設計パラメータ群候補を含む複数の前記設計パラメータ群候補を所定のアルゴリズムにより選択する、
請求項1~8のいずれか一項に記載の設計支援装置。 - 前記獲得関数構築部は、LCB(Lower Confidence Bound)、EI(Expected Improvement)及びPI(Probability of Improvement)のうちのいずれかからなる前記獲得関数を構築する、
請求項1~9のいずれか一項に記載の設計支援装置。 - 前記獲得関数構築部は、前記設計パラメータ群に応じて発生する、前記製品、前記仕掛品、前記半製品、前記部品又は前記試作品の作製に係る時間及び費用のうちの少なくともいずれかを含むコストに関するコスト値を含み、該コスト値が大きいほど、前記設計パラメータ群の好適の程度が減ぜられたことを示す前記指標値を出力する前記獲得関数を構築する、
請求項1~10のいずれか一項に記載の設計支援装置。 - 複数の設計パラメータからなる設計パラメータ群に基づいて作製される製品、仕掛品、半製品、部品又は試作品の設計において、設計パラメータの決定と決定された設計パラメータに基づく製品、仕掛品、半製品、部品又は試作品の作製との繰り返しにより設計パラメータの最適化を図る手法に適用するために、製品、仕掛品、半製品、部品又は試作品の特性を示す複数の特性項目のそれぞれについて設定された目標値を満たすような、前記複数の設計パラメータを求める設計支援装置における設計支援方法であって、
作製済みの前記製品、前記仕掛品、前記半製品、前記部品又は前記試作品に関しての、前記設計パラメータ群と前記複数の特性項目のそれぞれの観測値とからなる実績データを複数取得するデータ取得ステップと、
前記設計パラメータ群に基づいて前記特性項目の観測値を確率分布若しくはその近似又は代替指標として予測する予測モデルを、前記実績データに基づいて構築するモデル構築ステップと、
少なくとも前記予測モデルに基づいて、前記設計パラメータ群を入力とし前記特性項目に示される特性の向上に関する前記設計パラメータ群の指標値を出力とする獲得関数を前記特性項目ごとに構築する獲得関数構築ステップと、
前記複数の獲得関数のそれぞれの出力を目的変数とする前記設計パラメータ群についての多目的最適化により設計パラメータ群候補を複数生成する設計パラメータ群候補生成ステップと、
前記設計パラメータ群候補を前記予測モデルに入力することにより得られる前記観測値の確率分布若しくはその近似又は代替指標に基づいて、全ての特性項目の前記目標値が達成される確率である全体達成確率を前記設計パラメータ群候補ごとに算出し、前記全体達成確率が最も高い少なくとも一つの前記設計パラメータ群候補を選択する選択ステップと、
選択した前記設計パラメータ群候補を出力する出力ステップと、
を有する設計支援方法。 - コンピュータを、複数の設計パラメータからなる設計パラメータ群に基づいて作製される製品、仕掛品、半製品、部品又は試作品の設計において、設計パラメータの決定と決定された設計パラメータに基づく製品、仕掛品、半製品、部品又は試作品の作製との繰り返しにより設計パラメータの最適化を図る手法に適用するために、製品、仕掛品、半製品、部品又は試作品の特性を示す複数の特性項目のそれぞれについて設定された目標値を満たすような、前記複数の設計パラメータを求める設計支援装置として機能させるための設計支援プログラムであって、
前記コンピュータに、
作製済みの前記製品、前記仕掛品、前記半製品、前記部品又は前記試作品に関しての、前記設計パラメータ群と前記複数の特性項目のそれぞれの観測値とからなる実績データを複数取得するデータ取得機能と、
前記設計パラメータ群に基づいて前記特性項目の観測値を確率分布若しくはその近似又は代替指標として予測する予測モデルを、前記実績データに基づいて構築するモデル構築機能と、
少なくとも前記予測モデルに基づいて、前記設計パラメータ群を入力とし前記特性項目に示される特性の向上に関する前記設計パラメータ群の指標値を出力とする獲得関数を前記特性項目ごとに構築する獲得関数構築機能と、
前記複数の獲得関数のそれぞれの出力を目的変数とする前記設計パラメータ群についての多目的最適化により設計パラメータ群候補を複数生成する設計パラメータ群候補生成機能と、
前記設計パラメータ群候補を前記予測モデルに入力することにより得られる前記観測値の確率分布若しくはその近似又は代替指標に基づいて、全ての特性項目の前記目標値が達成される確率である全体達成確率を前記設計パラメータ群候補ごとに算出し、前記全体達成確率が最も高い少なくとも一つの前記設計パラメータ群候補を選択する選択機能と、
選択した前記設計パラメータ群候補を出力する出力機能と、
を実現させる設計支援プログラム。
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US18/256,450 US20240028795A1 (en) | 2020-12-10 | 2021-11-24 | Design assitance device, design assitance method, and design assitance program |
| CN202180082638.2A CN116601635A (zh) | 2020-12-10 | 2021-11-24 | 设计支援装置、设计支援方法及设计支援程序 |
| EP21903179.6A EP4246363A4 (en) | 2020-12-10 | 2021-11-24 | Design assitance device, design assitance method, and design assitance program |
| KR1020237020668A KR20230113571A (ko) | 2020-12-10 | 2021-11-24 | 설계 지원 장치, 설계 지원 방법 및 설계 지원 프로그램 |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2020-205027 | 2020-12-10 | ||
| JP2020205027A JP7604871B2 (ja) | 2020-12-10 | 2020-12-10 | 設計支援装置、設計支援方法及び設計支援プログラム |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2022124075A1 true WO2022124075A1 (ja) | 2022-06-16 |
Family
ID=81972855
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2021/042983 Ceased WO2022124075A1 (ja) | 2020-12-10 | 2021-11-24 | 設計支援装置、設計支援方法及び設計支援プログラム |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US20240028795A1 (ja) |
| EP (1) | EP4246363A4 (ja) |
| JP (1) | JP7604871B2 (ja) |
| KR (1) | KR20230113571A (ja) |
| CN (1) | CN116601635A (ja) |
| WO (1) | WO2022124075A1 (ja) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024024957A1 (ja) * | 2022-07-29 | 2024-02-01 | 株式会社レゾナック | 設計支援装置、設計支援方法及び設計支援プログラム |
| CN117574721A (zh) * | 2023-11-20 | 2024-02-20 | 华中科技大学 | 一种工艺参数概率模型优化方法 |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP7310735B2 (ja) * | 2020-07-01 | 2023-07-19 | トヨタ自動車株式会社 | 多性能最適化設計装置、及び多性能最適化設計方法 |
| WO2024070169A1 (ja) * | 2022-09-29 | 2024-04-04 | 日本碍子株式会社 | 試作条件提案システム、試作条件提案方法 |
| DE112023004307T5 (de) * | 2022-09-29 | 2025-08-14 | Ngk Insulators, Ltd. | Testproduktionsbedingungsvorschlagssystem und testproduktionsbedingungvorschlagsverfahren |
| JP2024172655A (ja) * | 2023-05-31 | 2024-12-12 | 株式会社日立製作所 | 設計支援システム、及び設計支援方法 |
| CN120188163A (zh) * | 2023-10-19 | 2025-06-20 | 株式会社力森诺科 | 设计支援装置、设计支援方法及设计支援程序 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014006825A (ja) * | 2012-06-27 | 2014-01-16 | Hitachi Ltd | 設計支援装置 |
| JP2016045536A (ja) * | 2014-08-20 | 2016-04-04 | 株式会社日立製作所 | 設計支援装置 |
| JP2016200902A (ja) * | 2015-04-08 | 2016-12-01 | 横浜ゴム株式会社 | 構造体の近似モデル作成方法、構造体の近似モデル作成装置、およびプログラム |
| WO2019088185A1 (ja) * | 2017-11-01 | 2019-05-09 | 株式会社日立製作所 | 設計支援装置及び設計支援方法 |
| JP2020052737A (ja) | 2018-09-27 | 2020-04-02 | 株式会社神戸製鋼所 | 製品設計装置および該方法 |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2913743C (en) * | 2013-05-30 | 2023-01-03 | Universite De Sherbrooke | Systems and methods for performing bayesian optimization |
| AU2020101453A4 (en) * | 2020-07-23 | 2020-08-27 | China Communications Construction Co., Ltd. | An Intelligent Optimization Method of Durable Concrete Mix Proportion Based on Data mining |
| JP7661693B2 (ja) * | 2020-12-10 | 2025-04-15 | 株式会社レゾナック | 設計支援装置、設計支援方法及び設計支援プログラム |
-
2020
- 2020-12-10 JP JP2020205027A patent/JP7604871B2/ja active Active
-
2021
- 2021-11-24 KR KR1020237020668A patent/KR20230113571A/ko active Pending
- 2021-11-24 WO PCT/JP2021/042983 patent/WO2022124075A1/ja not_active Ceased
- 2021-11-24 US US18/256,450 patent/US20240028795A1/en active Pending
- 2021-11-24 EP EP21903179.6A patent/EP4246363A4/en active Pending
- 2021-11-24 CN CN202180082638.2A patent/CN116601635A/zh active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014006825A (ja) * | 2012-06-27 | 2014-01-16 | Hitachi Ltd | 設計支援装置 |
| JP2016045536A (ja) * | 2014-08-20 | 2016-04-04 | 株式会社日立製作所 | 設計支援装置 |
| JP2016200902A (ja) * | 2015-04-08 | 2016-12-01 | 横浜ゴム株式会社 | 構造体の近似モデル作成方法、構造体の近似モデル作成装置、およびプログラム |
| WO2019088185A1 (ja) * | 2017-11-01 | 2019-05-09 | 株式会社日立製作所 | 設計支援装置及び設計支援方法 |
| JP2020052737A (ja) | 2018-09-27 | 2020-04-02 | 株式会社神戸製鋼所 | 製品設計装置および該方法 |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP4246363A4 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024024957A1 (ja) * | 2022-07-29 | 2024-02-01 | 株式会社レゾナック | 設計支援装置、設計支援方法及び設計支援プログラム |
| CN117574721A (zh) * | 2023-11-20 | 2024-02-20 | 华中科技大学 | 一种工艺参数概率模型优化方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| US20240028795A1 (en) | 2024-01-25 |
| CN116601635A (zh) | 2023-08-15 |
| EP4246363A1 (en) | 2023-09-20 |
| JP2022092297A (ja) | 2022-06-22 |
| EP4246363A4 (en) | 2024-05-15 |
| JP7604871B2 (ja) | 2024-12-24 |
| KR20230113571A (ko) | 2023-07-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022124075A1 (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| WO2022124077A1 (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| JP7676912B2 (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| Crampin et al. | Mathematical and computational techniques to deduce complex biochemical reaction mechanisms | |
| Styczynski et al. | Overview of computational methods for the inference of gene regulatory networks | |
| EP4097729A1 (en) | System and method for learning to generate chemical compounds with desired properties | |
| Cavalcanti et al. | Production system efficiency optimization using sensor data, machine learning-based simulation and genetic algorithms | |
| Dimitrova et al. | Parameter estimation for Boolean models of biological networks | |
| EP4535216A1 (en) | Design assistance device, design assistance method, and design assistance program | |
| JP7711560B2 (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| CN118898229A (zh) | 面向vlsi划分的切割平面方法、装置、计算机设备、及介质 | |
| EP4582995A1 (en) | Design support device, design support method, and design support program | |
| JP2024011665A (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| Sardeshmukh et al. | Bayesian networks for inverse inference in manufacturing Bayesian networks | |
| JP2025030855A (ja) | 設計支援装置、設計支援方法及び設計支援プログラム | |
| Hao et al. | Semiparametric Identification of the Discount Factor and Payoff Function in Dynamic Discrete Choice Models | |
| Pyle et al. | Surrogate modelling for efficient discovery of emergent population dynamics | |
| Lai et al. | Mania: A Gene Network Reverse Algorithm For Compounds Mode-Of-Action And Genes Interactions Inference | |
| Zhou et al. | 15 Reconstruction of | |
| CN119200434A (zh) | 一种控制数智化仿真系统及方法 | |
| Lai et al. | MANIA: A Gene Network Reverse Algorithm for Compounds Mode-of-Action and Genes Interactions Inference | |
| Johnson et al. | Supplemental Digital Content A new severity of illness scale using a subset of APACHE data elements shows comparable predictive accuracy | |
| Zhou et al. | 15 Reconstruction of Regulatory Networks from Microarray Data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21903179 Country of ref document: EP Kind code of ref document: A1 |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 202180082638.2 Country of ref document: CN Ref document number: 18256450 Country of ref document: US |
|
| ENP | Entry into the national phase |
Ref document number: 2021903179 Country of ref document: EP Effective date: 20230614 |
|
| ENP | Entry into the national phase |
Ref document number: 20237020668 Country of ref document: KR Kind code of ref document: A |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |