DESIGN SUPPORT METHOD, DESIGN SUPPORT PROGRAM, AND COMPUTER-READABLE STORAGE MEDIUM CONTAINING THE DESIGN SUPPORT PROGRAM
The design support method simplifies and speeds up the formulation of structure performance by using multiple extraction methods and genetic operations to select and combine performance factors, addressing computational inefficiencies in existing methods.
Patent Information
- Application Number
- JP2021033316
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-03-03
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-03-03
AI Technical Summary
Existing design support methods face complexity and computational inefficiencies due to the need to consider an infinite number of performance factors, including interactions and trigonometric functions, making it difficult to formulate the performance of structures quickly and simply.
A design support method that uses a computer system to formulate performance by selecting and combining performance factors through multiple extraction methods, setting a selection order based on extraction results, and applying genetic operations on a tree structure to narrow down candidates, while excluding irrelevant design variables and using sparse and stepwise regression analyses.
This approach allows for the formulation of structure performance simply and quickly, accurately capturing important interactions and reducing computational burden, enabling efficient design support.
Smart Images

Figure 0007589592000006 
Figure 0007589592000007 
Figure 0007589592000008
Abstract
Description
[Technical field]
[0001] The technology disclosed herein relates to a design support method, a design support program, and a computer-readable storage medium storing the design support program. [Background technology]
[0002] For example, Patent Document 1 discloses a design support method for a structure having a plurality of parts. Specifically, the design support method for a structure disclosed in Patent Document 1 is configured to set a linear combination of design variables obtained by quantifying the specifications of each part as a response variable, and to identify a model indicating a performance index of the structure by performing a multiple regression analysis on the response variable. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2020-71725 A Summary of the Invention [Problem to be solved by the invention]
[0004] Incidentally, when formulating the performance of a structure using design variables as disclosed in Patent Document 1, in order to take into account more complex phenomena such as the mutual influences that parts have on each other, it is conceivable to formulate the performance by regarding each term expressed in a more general functional form, such as an interaction term obtained by multiplying design variables together, or a term expressed by a trigonometric function with the design variables as an argument, as a performance factor.
[0005] However, since there are an infinite number of such performance factors, the calculations required to formulate the model become complicated and the time required for the calculations may become enormous, which is inconvenient for smoothly carrying out design support using the model.
[0006] The technology disclosed herein has been made in consideration of the above points, and its purpose is to formulate the performance of a structure simply and quickly. [Means for solving the problem]
[0007] A first aspect of the present disclosure relates to a design support method for formulating the performance of a structure composed of multiple parts using design variables that quantify the specifications of each part, by using a computer having a calculation unit that executes a program and a storage unit that reads data.
[0008] According to a first aspect of the present disclosure, the design support method includes a reading step in which the memory unit reads in a set of performance factors each having one or more of the design variables as arguments; an extraction step in which the calculation unit applies a plurality of types of extraction methods to the set of performance factors to extract performance factors that contribute to the performance of the structure for each extraction method; a ranking setting step in which the calculation unit sets a selection ranking of the performance factors based on the extraction results obtained for each extraction method; and a formulation step in which the calculation unit selects the performance factors from the set of performance factors in accordance with the selection ranking, and formulates the performance of the structure by combining the selected performance factors.
[0009] Here, the "performance factors" include factors used in polynomial approximation of design variables, such as first-order terms of design variables and second-order terms of design variables, as well as factors that realize non-polynomial approximation, such as trigonometric functions with design variables as arguments.
[0010] According to the first aspect, the design support method applies a plurality of types of extraction methods to a set of performance factors to set a selection order for the performance factors. This selection order can be said to indicate the importance of the performance factors. When setting the selection order, by using a plurality of extraction methods individually, rather than using only a specific extraction method, the selection order can be set more appropriately. Then, performance factors with a relatively high selection order are actively incorporated into a model formula describing the performance, while performance factors with a relatively low selection order are passively incorporated into the model formula, thereby making it possible to faithfully describe the performance of the structure while suppressing the number of terms used in formulation. This makes it possible to formulate the performance of the structure simply and quickly.
[0011] According to a second aspect of the present disclosure, the performance factor may include a linear term formed of the design variables themselves, and a non-linear term formed by multiplying the design variables together.
[0012] According to the second aspect, the performance factor includes a nonlinear term obtained by multiplying the design variables together. This allows the model to include interactions related to the connections between parts. This is effective in modeling dynamic phenomena such as the crash performance of a structure.
[0013] In addition, according to a third aspect of the present disclosure, the design support method may include a pre-narrowing step in which the calculation unit sets a set of performance factors prior to the extraction step, and in the pre-narrowing step, the set of performance factors is set by performing a genetic operation on a tree structure in which the design variables are terminal nodes and multiplication operators are non-terminal nodes, and in the pre-narrowing step, the genetic operation is performed so that the number of levels of the tree structure is equal to or less than a predetermined maximum depth.
[0014] According to the third aspect, prior to the selection of the performance factors based on the selection order, the performance factors are narrowed down using a tree structure. Here, by excluding the coefficients of the design variables from the terminal node, the narrowing down can be performed more quickly. At that time, by setting the number of levels of the tree structure in advance, the narrowing down can be performed suitable for the structure to be supported. This allows for more efficient formulation.
[0015] In addition, according to a fourth aspect of the present disclosure, the design support method may include a model reconstruction step in which the calculation unit excludes from the performance of the structure the nonlinear terms consisting of design variables that do not appear in the performance of the structure as the linear terms, among the design variables that constitute the performance factor.
[0016] According to the fourth aspect, among the design variables constituting the nonlinear terms, design variables that do not appear in the model as first-order linear terms (for example, design variables that do not appear in the model as first-order terms (main effects) but appear in the model only as interaction terms) are excluded from the model. By carrying out this step, the number of performance factors to be considered can be reduced, and therefore the amount of calculation can be reduced.
[0017] In addition, according to a fifth aspect of the present disclosure, the calculation unit may determine an extraction frequency for each of the performance factors based on the extraction results obtained for each of the extraction methods, and the calculation unit may set the selection order based on the level of the extraction frequency.
[0018] According to the fifth aspect, a performance factor with a high extraction frequency, for example, a performance factor extracted by all extraction methods, is set to a high selection order, while a performance factor with a low extraction frequency, for example, a performance factor extracted by only one of the extraction methods, is set to a low selection order. By setting in this way, the selection order can be set more appropriately, and the performance of the structure can be formulated more faithfully.
[0019] According to a sixth aspect of the present disclosure, the performance of the structure is formulated by a linear combination of the performance factors, and the calculation unit executes a multiple regression analysis for each of the plurality of types of extraction methods, with the performance factors as explanatory variables and the performance of the structure as a target variable, and the calculation unit determines the extraction frequency based on the coefficients of each of the explanatory variables calculated by the multiple regression analysis.
[0020] According to the sixth aspect, the calculation unit determines the extraction frequency based on the magnitude of the coefficient multiplied by each performance factor. In this case, for example, a performance factor whose coefficient is calculated as zero in a specific extraction method can be determined as non-extracted at least in that extraction method. By setting in this way, the extraction frequency can be clearly determined, which is advantageous in formulating the performance of a structure simply and quickly.
[0021] According to a seventh aspect of the present disclosure, the plurality of types of extraction methods may include a regression analysis having sparsity.
[0022] According to the seventh aspect, when a regression analysis with sparsity is performed, coefficients with low contribution are set to zero. This makes it possible to clearly determine the extraction frequency, which is advantageous in formulating the performance of a structure simply and quickly.
[0023] According to an eighth aspect of the present disclosure, the plurality of types of extraction methods may include a stepwise regression analysis.
[0024] According to the eighth aspect, the coefficients of the performance factors that were not selected when the stepwise method was performed can be regarded as zero. This makes it possible to clearly determine the extraction frequency, which is advantageous in formulating the performance of the structure simply and quickly.
[0025] According to a ninth aspect of the present disclosure, the design support method may include an index calculation step of calculating an evaluation index characterizing an accuracy of the formulation in the formulation step, and a verification step of verifying the accuracy of the formulation in the formulation step based on the evaluation index calculated by the index calculation step, wherein the storage unit reads in a plurality of sample data each corresponding to a set of performance factors, and the calculation unit divides the plurality of sample data into learning data and verification data according to a predetermined division pattern, and the calculation unit executes the extraction step, the rank setting step, and the formulation step based on the learning data, while executing the index calculation step based on the remaining verification data, and the calculation unit changes the division pattern a plurality of times, and repeatedly executes the extraction step, the rank setting step, the formulation step, and the index calculation step for each changed division pattern, and then executes the verification step, and in the verification step, the calculation unit selects any one of the performances of the structure formulated for each division pattern based on the evaluation index calculated for each division pattern.
[0026] According to the ninth aspect, in the design support method, k-fold cross-validation is performed when formulating based on the selection ranking. That is, formulation is performed based on k types of learning data, and an evaluation index is calculated based on the remaining validation data for each formulation. Then, by selecting the model with the best evaluation index from among the k types of models obtained by each division pattern, it becomes possible to perform a more robust formulation. This is effective in achieving both a simple and fast formulation of the performance of the structure and robustness of the model formula.
[0027] According to a tenth aspect of the present disclosure, the structure may be a vehicle body, and the design variable may be a plate thickness of each of the plurality of components.
[0028] An automobile body is composed of many parts. In this case, it is considered that many performance factors need to be considered. The present disclosure is effective in that it can actively narrow down important performance factors from among many performance factors in a technical field in which many performance factors can be considered but it is not easy to secure sample data beyond that.
[0029] An eleventh aspect of the present disclosure relates to a design support program that, when executed by a computer having a calculation unit that executes a program and a memory unit that reads data, formulates the performance of a structure composed of multiple parts using design variables that quantify the specifications of each part.
[0030] According to an eleventh aspect of the present disclosure, the design support program causes the computer to execute the following steps: a reading step in which the memory unit reads in a set of performance factors with one or more of the design variables as arguments; an extraction step in which the calculation unit applies a plurality of types of extraction methods to the set of performance factors to extract performance factors that contribute to the performance of the structure for each extraction method; a ranking setting step in which the calculation unit sets a selection ranking of the performance factors based on the extraction results obtained for each extraction method; and a formulation step in which the calculation unit selects the performance factors from the set of performance factors in accordance with the selection ranking, and formulates the performance of the structure by combining the selected performance factors.
[0031] According to the eleventh aspect, the performance of a structure can be formulated simply and quickly.
[0032] A twelfth aspect of the present disclosure relates to a computer-readable storage medium. The storage medium stores the design support program according to the eleventh aspect. Effect of the Invention
[0033] As described above, according to the present disclosure, the performance of a structure can be formulated simply and quickly. [Brief description of the drawings]
[0034] [Figure 1] FIG. 1 is a diagram illustrating a hardware configuration of a design support apparatus. [Diagram 2] FIG. 2 is a diagram illustrating an example of a software configuration of the design support apparatus. [Diagram 3] FIG. 3 is a flowchart illustrating the procedure of the design support method. [Figure 4] FIG. 4 is a flowchart illustrating a specific procedure of the preliminary narrowing down phase. [Diagram 5] FIG. 5 is a diagram for explaining the relationship between a tree structure used in the preliminary narrowing down phase and a model formula corresponding to the tree structure. [Figure 6] FIG. 6 is a diagram for explaining the one-point crossover method in the preliminary narrowing down phase. [Figure 7] FIG. 7 is a flowchart illustrating a specific procedure of the performance factor extraction phase. [Figure 8] FIG. 8 is a diagram for explaining the procedure for acquiring the performance factors according to the selection order. [Figure 9] FIG. 9 is a diagram for explaining the k-fold intersection method. [Figure 10] FIG. 10 is a diagram showing a specific example of the performance factor extraction phase. [Figure 11] FIG. 11 is a diagram illustrating the influence of the interaction terms on the crash performance values. [Figure 12] FIG. 12 is a diagram for explaining the model reconstruction step. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0035] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. Note that the following description is merely an example.
[0036] <Device configuration> Fig. 1 is a diagram illustrating an example of a hardware configuration of a design support apparatus (specifically, a computer 1 constituting the design support apparatus) according to the present disclosure, and Fig. 2 is a diagram illustrating an example of a software configuration of the design support apparatus.
[0037] 1, the computer 1 includes a central processing unit (CPU) 3 that controls the entire computer 1, a read only memory (ROM) 5 that stores a boot program and the like, a random access memory (RAM) 7 that functions as a main memory, and a hard disk drive (HDD) 9 as a secondary storage device. Note that instead of the HDD 9, a solid state drive (SSD) or the like can also be used as the secondary storage device.
[0038] Of these elements, the CPU 3 executes various programs. The CPU 3 functions as a calculation unit in this embodiment. The RAM 7 temporarily stores the programs executed by the CPU 3 and various data, and the HDD 9 continuously stores the programs, data, and the like. The RAM 7 can also read data from the HDD 9 as necessary. The RAM 7 functions as a storage unit in this embodiment.
[0039] The computer 1 also includes a display unit 11 consisting of a liquid crystal display, an organic EL display, a CRT display, or the like, a graphics memory (Video RAM: VRAM) 13 for storing image data to be displayed on the display unit 11, and a keyboard 15 and a mouse 17 as man-machine interfaces. The display unit 11 can display the results of calculations performed by the CPU 3. The computer 1 according to this embodiment can also transmit and receive data to and from external devices via a communication interface 21.
[0040] As described below, the design support apparatus according to the present disclosure may be configured with a plurality of computers 1. In this case, the CPU 3 and RAM 7 of each computer 1 can be regarded as the arithmetic unit and storage unit according to the present disclosure, respectively.
[0041] As illustrated in FIG. 2, the program memory of the HDD 9 stores an operating system (OS) 19, a pre-narrowing program 29A, a performance factor extraction program 29B, a design knowledge extraction program 29C, and the like.
[0042] Among these elements, the preliminary narrowing down program 29A, the performance factor extraction program 29B, and the design knowledge extraction program 29C constitute a design support program 29 in this embodiment.
[0043] Here, the design support program 29 is a program configured to cause the computer 1 to execute the design support method according to this embodiment, and can cause the computer 1 to execute each step constituting the design support method. The design support program 29 is pre-stored in the computer-readable storage medium 18 (see FIG. 1).
[0044] In the program memory of the HDD 9, the pre-narrowing program 29A, the performance factor extraction program 29B, and the design knowledge extraction program 29C are each started in response to an instruction inputted from the keyboard 15, the mouse 17, etc. At that time, the pre-narrowing program 29A, etc. are loaded from the HDD 9 to the RAM 7 and executed by the CPU 3. The CPU 3 executes the pre-narrowing program 29A, etc., so that the computer 1 functions as a design support device.
[0045] Meanwhile, the data memory of the HDD 9 stores a plurality of sample data 49. Each sample data 49 is defined as multivariate data consisting of a plurality of independent variables. Each sample data 49 is loaded from the HDD 9 to the RAM 7 as necessary and used for calculations by the CPU 3.
[0046] Here, each sample data 49 indicates the specifications of a structure to be analyzed and the performance values realized by the structure. This structure is composed of multiple parts. Each sample data 49 also includes multiple design variables x i (i=1, 2, ..., N) and a response y (also called performance value y or objective variable y) output when the specification is input are stored. Each sample data 49 may be generated in step S12, step S21, etc. described later, or may be generated in advance and read from outside and stored.
[0047] For example, when designing an automobile body using model-based development, the structure corresponds to the vehicle body, and the multiple parts correspond to vehicle body components such as B-pillars, side sills, front tunnels, roof rails, front frames, side frames, rear side frames, cross members, floor panels, front panels, rear panels, engine covers, bonnets, rear fenders, roofs, doors, lift gates, rear end panels, rear wheel house outers, rear end members, rear pillar outers, rear pillar inners, rear pillar reinforcements, upper reinforcements, and rear roof rails. In this case, the design variables x corresponding to each part are i The plate thickness, shape, joint strength, weight, material, etc. of the corresponding part can be used as the response y. In this case, the collision performance value of the vehicle body (a performance index indicating the resistance to deformation during a collision), the torsional rigidity of the vehicle body (a performance index indicating the resistance to torsional deformation), the driving stability, the NVH performance, etc. can be used as the response y.
[0048] In addition, various data generated by executing the pre-narrowing program 29A, the performance factor extraction program 29B, and the design knowledge extraction program 29C, as well as the execution results of the application program 39, are stored in the data memory of the HDD 9 or temporarily stored in the RAM 7 serving as the main memory, as necessary.
[0049] The specific methodology of the design support method will be described in detail below.
[0050] <Methodology> Fig. 3 is a flow chart illustrating the procedure of a design support method. The flow illustrated in Fig. 3 is a design support method in which the performance of a structure composed of multiple parts is formulated by using design variables in which the specifications of each part are quantified, by using a computer 1.
[0051] The design support method is basically implemented by sequentially executing a pre-narrowing phase S1 as a pre-narrowing step, a performance factor extraction phase S2 consisting of a reading step, an extraction step, a ranking step, a formulation step, an index calculation step, and a verification step, and a design knowledge extraction phase S3 consisting of a model reconstruction step. Note that the pre-narrowing phase S1 and the design knowledge extraction phase S3 are not essential.
[0052] Among these phases S1 to S3, the pre-narrowing phase S1 is implemented by the CPU 3 executing the above-mentioned pre-narrowing program 29A. Similarly, the performance factor extraction phase S2 is implemented by the CPU 3 executing the performance factor extraction program 29B, and the design knowledge extraction phase S3 is implemented by the CPU 3 executing the design knowledge extraction program 29C.
[0053] Each phase constituting the design support method will be described below in order.
[0054] (Pre-narrowing phase S1) Fig. 4 is a flow chart illustrating a specific procedure of the pre-narrowing phase S1. Fig. 5 is a diagram for explaining the relationship between a tree structure T1 used in the pre-narrowing phase S1 and a response y corresponding to the tree structure T1. Fig. 6 is a diagram for explaining the one-point crossover method in the pre-narrowing phase S1.
[0055] In this embodiment, the performance factor X j (j=1,2,…,M) Formulate the performance of the structure by a linear combination, where the performance factor X j The design variable x i Then, each performance factor X j Contribution of work to β j The term β multiplied by j X j By calculating a linear combination of these, the response y is formulated as shown in the lower part of FIG. 5. In the example shown in FIG. 5, the response y is formulated by a linear combination of the intercept and four performance factors X1 to X4. Each performance factor X j is the design variable x i where the performance factor X j is the design variable x i In addition to the term multiplied by one or more of the design variables x i Trigonometric function with arguments, design variable x i n-th root of design variable x i Logarithm of design variable x i This includes exponential functions that take
[0056] Here, to formulate the response y, the performance factor X that contributes to the performance of the structure is j and extracting each performance factor X j Contribution factor β to be multiplied j In particular, when considering nonlinear terms such as interactions described below, the design variable x i If there are many, the performance factor X j There are thought to be countless variations.
[0057] The inventors of the present application have determined that the performance factor Xj and contribution rate β j Prior to the performance factor extraction phase S2 for determining the performance factor X j We came up with the idea of carrying out a process to narrow down the combinations in advance (pre-narrowing phase S1).
[0058] That is, the pre-narrowing phase S1 according to this embodiment is a process executed prior to the performance factor extraction phase S2, in which the CPU 3 selects the performance factor X j This is a process for setting a set (candidate list).
[0059] Specifically, in this embodiment, the CPU 3 executes genetic programming (GP) to generate N design variables x i The response y is formulated from the above. As is well known, GP is a genetic algorithm (GA) that has been expanded to allow it to be expressed in a tree structure. In the commonly known GP, the performance factor X j and the contribution rate β to be multiplied by it j and are set as the terminal nodes, respectively, to determine both simultaneously.
[0060] In contrast, in the pre-narrowing phase S1 according to the present embodiment, M performance factors X j Determine the set of performance factors X j Contribution factor β to be multiplied j are determined in the subsequent performance factor extraction phase S2. In other words, the tree structure used in the preliminary refinement phase S1 is i In the example shown in Figure 5, four performance factors x1, x3, x1 x2, x3 are described based on three design variables x1, x2, and x3. 2 is now being determined.
[0061] Specifically, first, in step S11, a maximum depth is set. This maximum depth is a parameter that is referenced when performing one-point crossover in step S17 described later.
[0062] In the next step S12, first, N design variables x i Based on the performance factor X j The initial set of combinations (initial population) of design variables x is generated widely throughout the entire design space. i For example, the number of design variables x i The number can be 1 to 4 times the number of
[0063] In the next step S13, a list set (performance factor X j For example, when step S13 is executed for the first time, the design variable x i In this case, each term of the L-th order polynomial is approximated by a performance factor X j Then, by performing multiple regression analysis with reference to sample data 49, the contribution ratio β j Then, the initial population is evaluated by evaluating the mean square error for the determined polynomial.
[0064] In the following step S14, it is determined whether the mean squared error calculated in step S13 has converged, and if it has converged (step S14: YES), the process proceeds to step S15, where a candidate list, which is the set of lists at the time of convergence, is output.
[0065] On the other hand, if convergence is not achieved in step S14 (step S14: NO), the process proceeds to step S16 to select an excellent list set. Here, several parent individuals with high fitness and distance from each other are selected from the current generation. Then, in the following step S17, one-point crossover, which is equivalent to gene recombination, is performed.
[0066] Specifically, in steps S17 and S18, the design variable x i By performing genetic operations on a tree structure with j Set a set of .
[0067] Specifically, in step S17, as shown in FIG. 6, one-point crossover is performed between the first list set L1 and the second list set L2 to generate new list sets L1' and L2'. At that time, as can be seen from FIG. 6, the contribution degree β j is not a terminal node, and the performance factor X j In the example shown, the fourth performance factor N1 (=x3 2 ) and the fourth performance factor N1(=x2 2 ) and are subject to recombination.
[0068] Here, the genetic operation is executed so that the number of layers of the tree structure is equal to or less than the predetermined depth (maximum depth) set in step S11. For example, if the maximum depth is set to 2 and the multiplication operator is set as a non-terminal node as described above, the performance factor X j contains the design variable x i A linear term of the first order consisting of itself and the design variable x i In this case, the latter nonlinear term includes a specific design variable x i The term x obtained by squaring i 2 and different design variables x i , x j The so-called interaction term x i x j The following will be included:
[0069] In step S18, which follows step S17, a sudden mutation is performed to stochastically change the genetic information to give diversity to the individual, and an individual of a child generation is generated. Then, steps S13 to S18 are repeated, and when the optimal solution is converged to or a predetermined number of evolutionary generations is reached, the process proceeds from step S14 to step S15, and the calculation ends. Here, the predetermined number of evolutionary generations can be determined according to the number of design variables, etc., and can be, for example, 100 generations to 500 generations.
[0070] Although not shown in detail, k-fold cross-validation may be performed in the pre-narrowing phase S1. In this case, the sample data 49 referred to in the multiple regression analysis may be divided into learning data and validation data.
[0071] It should be noted that a sequential approximation optimization method may be used to address problems such as increased calculation time and over-learning that accompany an increase in the number of design variables. The sequential approximation optimization method is a method in which an approximation equation is found from a relatively small number of sample points and the response value of the objective function using a response surface, and an optimal solution is searched for using this, and the optimal solution and some sample points are added. The sequential approximation optimization method can reduce the amount of calculation per CAE, and improve the efficiency of optimization calculations.
[0072] When the control process proceeds to step S15, the flow shown in FIG. 4 ends, and the control process proceeds from the pre-narrowing phase S1 to the performance factor extraction phase S2 in FIG.
[0073] (Performance factor extraction phase S2) FIG. 7 is a flow chart illustrating a specific procedure of the performance factor extraction phase S2. FIG. j FIG. 2 is a diagram for explaining the import procedure of FIG. 9 is a diagram showing a specific example of the performance factor extraction phase S2.
[0074] First, in step S21 as a reading step, the RAM 7 as a storage unit reads the performance factor X j In step S21, the CPU 3 reads in the set of design variables x i The values of performance factors X j By substituting the result of the substitution with the response y, the performance factor X j A large number of sample data Dt corresponding to the set of
[0075] In the following step S22, the RAM 7 reads a plurality of sample data Dt, and the CPU 3 divides the plurality of sample data Dt into learning data Ds and verification data Dv in accordance with predetermined division patterns P1 to P9.
[0076] Specifically, in step S22, the CPU 3 divides the sample data Dt into k data blocks to perform k-fold cross validation. Then, as shown in Fig. 9, among each of the k divided data blocks, one data block is set as verification data Dv, and the remaining (k-1) data blocks are set as learning data Ds. That is, there are k different settings of the verification data Dv and the learning data Ds in total. Hereinafter, the k different settings are referred to as division patterns P1 to Pk.
[0077] Based on the learning data Ds, the CPU 3 executes step S23 as an extraction step, step S24 as a ranking setting step, step S25 as a formulation step, and step S26, while executing step S27 as an index calculation step based on the remaining verification data Dt.
[0078] Next, in step S23 as an extraction step, the CPU 3 extracts a performance factor X j By applying multiple types of extraction methods to the set (list set), the performance factor X that contributes to the performance of the structure for each extraction method is calculated. j Extract.
[0079] The response y, which indicates the performance of a structure, is calculated by the function of one or more performance factors X, as described above. j The CPU 3 applies multiple extraction methods to the linear combination to obtain each performance factor X j A multiple regression analysis is performed using the above as explanatory variables and the response y represented by the linear combination as a response variable.
[0080] In this step S23, each performance factor X jBy performing a regression analysis using the value of as input and the value of the response y as the objective variable as output, the performance factor X j Contribution factor β, which is the coefficient to be multiplied by j is calculated. j The value of may vary for each extraction method applied.
[0081] Here, the extraction method applied in step S23 includes a regression analysis method having sparsity and a stepwise regression analysis method. Furthermore, as the former regression analysis method, for example, a lasso regression with L1 regularization (Lasso), an elastic net regression in which a ridge term is added to the L1 regularization, and an adaptive lasso regression in which a weight is added to the L1 regularization can be used. Note that all of these methods use a regularization parameter λ. The magnitude of the regularization parameter λ can be determined using a k-fold cross-validation method. In this case, in step S23, apart from the division of the sample data described in step S22, the learning data Ds is further divided into a plurality of data blocks in order to perform cross-validation for determining the regularization parameter λ. In this case, apart from the loop related to cross-validation executed in steps S22 to S28, a loop related to cross-validation is executed for each of the division patterns P1 to Pk in step S23. Here, in the cross-validation for determining the regularization parameter λ, an evaluation index without a penalty term, such as a deviance, is used. Therefore, a large number of contribution degrees β j becomes a non-zero value, which leaves a wide range of options available when narrowing down the options using the selection probabilities described below.
[0082] In addition, when using a regression analysis method with sparsity such as Lasso regression, the contribution rate β that contributes greatly to the objective variable j A significant value (non-zero value) is obtained for the performance factor X1, whereas a value of zero is obtained for the contribution factor βj that does not contribute relatively. j For the set of j It is possible to further narrow down (extract) the results based on the size of the results.
[0083] In addition, when applying the stepwise regression analysis method in step S23, the performance factor X can be calculated by using the information criteria AIC, BIC, etc. expressed by the following formula. j Among the set of performance factors that contribute to the performance of the structure, j will be extracted.
[0084]
number
[0085]
number
[0086] where n is the number of samples and σ 2 The variables with a hat in are the mean squared error (MSE) based on the estimator, and the variables with a hat in df are the degrees of freedom of the estimated model. AIC and BIC are used as model evaluation indices. The smaller the value of this evaluation indices, the better the model accuracy. Using this, the stepwise method estimates the performance factor X j Each element of the set is added to or removed from the linear combination one by one, and the AIC or BIC obtained at that time is judged to obtain the performance variable X j It is possible to further narrow down (extract) the set of performance factors. In addition, as a method for selecting and discarding performance factors in the stepwise method, an increase method, a decrease method, and an increase and decrease method can be considered, and any of these methods may be used. In the specific example described later, the increase method is configured to be used.
[0087] Next, in step S24 as a ranking step, the performance factors X obtained by the extraction methods are j Based on the extraction results, CPU3 has performance factor X j In detail, CPU3 selects the performance factor X obtained by each extraction method. jBased on the extraction results, performance factor X j The extraction frequency is determined for each item, and the selection order is set based on the extraction frequency.
[0088] For example, as illustrated in the upper table of FIG. 8, if the performance factor X1=x1 is extracted (≠0) in all of the three methods A to C, the extraction frequency of the performance factor can be determined to be high. On the other hand, if the performance factor X3=x3 in the third column is not extracted (0) in any of the methods A to C, the extraction frequency of the performance factor can be determined to be low. In the example illustrated in FIG. 8, the CPU 3 sets the selection order of the performance factor X1=x1 in the first column to the first with an extraction probability (selection) of 100%, the selection order of the performance factor X4=x4 in the fourth column to the second with a selection probability of 66%, the selection order of the performance factor X2=x2 in the second column to the third with a selection probability of 33%, and the selection order of the performance factor X3=x2 in the third column to the fourth with a selection probability of 0%. The selection order thus set may be temporarily stored in the RAM 7 or continuously stored in the HDD 9.
[0089] The frequency of occurrence is determined based on the performance factor X calculated in step S23. j Each coefficient (contribution rate β j ) can be used. As mentioned above, when using sparse regression analysis or stepwise regression analysis in multiple regression analysis, coefficients with low contribution become zero. Therefore, the contribution β j If is zero, the performance factor X j is considered not to have been extracted, and the contribution rate β j If is nonzero, the performance factor X j can be considered as having been extracted.
[0090] Here, when a more general method is used for multiple regression analysis, the coefficient with a low contribution does not necessarily become zero. In that case, the contribution β j If the value of exceeds a certain threshold, the performance factor X jis extracted, and if it is below the threshold, the performance factor X j In this case, it can be determined that the contribution degree β j You can prepare it separately, or use performance factor X. j If standardization is applied to the contribution ratio β j A common value may be used between the
[0091] Next, in step S25 as a formulation step, the CPU 3 calculates the performance factor X j From the set of j CPU3 also selects the performance factor X j The performance of the structure is formulated by combining these.
[0092] Specifically, in step S25, the performance factor X j The response y is formulated as the objective variable by linearly combining the above. In the example shown in Figure 8, the performance factor X1=x1, which has the first selection order, is incorporated first. In this case, the response y is a one-variable function of x1 that combines the intercept β0 and the linear term β1·x1 (first loop).
[0093] Then, in step S26 following step S25, the CPU 3 executes a multiple regression analysis on the objective variable formulated in step S25. In the first loop in Fig. 8, a multiple regression analysis using the learning data Ds is executed on the one-variable function formulated using x1. In the example shown in Fig. 8, a multiple regression analysis is executed on the one-variable function of x1, and an intercept β0 and a coefficient β1 of x1 are calculated. This completes the formulation of the response y in the first loop.
[0094] Next, in step S27 as an index calculation step, the CPU 3 calculates an evaluation index CV that characterizes the accuracy of the formulation executed in step S25 (more specifically, the accuracy of the model obtained by the formulation). l(l=1, 2, ..., l). The CPU 3 calculates the evaluation index CV l Based on the performance factor X j This determination is made by, for example, inputting the verification data Dv set in step S22 for the response y formulated in step S26, and calculating a predetermined evaluation index CV based on the output. l This can be done by calculating:
[0095] Specifically, in this embodiment, the evaluation index CV l In addition to the AIC and BIC mentioned above, the relaxed Cp criterion Cp mit can also be used.
[0096]
number
[0097] In addition, in this embodiment, the evaluation index CV l As the adjusted coefficient of determination R 2 and the root mean square error RMSE can be used.
[0098]
number
[0099]
number
[0100] In the above formula, the norm with the subscript “2” indicates the L2 norm.
[0101] For example, the evaluation index CV l If AIC is used as the coefficient, in step S27, the calculated AIC is compared with a predetermined threshold value, and if AIC is below the threshold value, the process proceeds to step S28, whereas if AIC is equal to or greater than the threshold value, the process returns to step S25.
[0102] Here, multiple evaluation indices CV l Alternatively, a plurality of evaluation indexes may be calculated in parallel, and the decision as to whether to proceed to step S28 or not may be made by majority vote.
[0103] Next, when returning from step S27 to step S25, the CPU 3 selects the previously selected performance factor X j Performance factor X ranked lower than j and linearly combines it with the response y to update the response y. In the example shown in FIG. 8, the performance factor X4=x4, which is second in the selection order, is incorporated second. In this case, the response y is a two-variable function combining the intercept β0, the linear term β1·x1, and the linear term β4·x4 (second loop). Thereafter, the CPU 3 advances the control process to step S26, and updates the contribution β j In the example shown in FIG. 8, multiple regression analysis is performed on the two-variable function of x1 and x4, and the intercept β0, the coefficient β1 of x1, and the coefficient β4 of x4 are calculated. This completes the formulation of the response y in the second loop. Thereafter, the CPU 3 proceeds to the control process S27, and updates the evaluation index CV l A judgment based on the above is made.
[0104] The CPU 3 repeats steps S25 to S27 to obtain a performance factor X j The objective variables are updated by taking in the above in order. l If the determination regarding the above is satisfied (for example, if the AIC is less than the threshold value), the CPU 3 selects the current division pattern P l The formulation of the response y at (l=1, 2, ..., k) is completed, and the control process proceeds from step S27 to step S28. At that time, the CPU 3 obtains the function form of the formulated response y and the evaluation index CV l The value of is temporarily stored in the RAM 7 or continuously stored in the HDD 9.
[0105] In step S28, the CPU 3 determines whether or not the end condition of the cross-validation is satisfied. Specifically, in this step, the CPU 3 performs the calculation of all the k division patterns P l If the answer is YES, the process proceeds to step S29. If the answer is NO, the division pattern P l After changing the above, the process returns to step S22.
[0106] In this manner, in this embodiment, the CPU 3 determines the division pattern P l is changed k times, and steps S23 to S28 are repeatedly executed for each changed division pattern P, and then the control process proceeds to step S29.
[0107] Next, in step S29, the CPU 3 calculates the evaluation index CV l In detail, in step S29, the CPU 3 verifies the accuracy of the formulation performed in step S25 based on the division pattern P l Evaluation index CV calculated for each l Based on the division pattern P l In more detail, the CPU 3 selects one of the k responses y formulated for each evaluation index CV l Compare them to find the smallest evaluation metric CV l The division pattern P corresponding to l、 In turn, the contribution rate β j , performance factor X j and response y. The CPU 3 selects the selected division pattern P l Contribution rate β corresponding to j etc. are output as the final result.
[0108] When step S29 is completed, the control process ends the flow shown in FIG. 7, and proceeds from the performance factor extraction phase S2 to the design knowledge extraction phase S3 in FIG.
[0109] (Design knowledge extraction phase S3) FIG. 12 is a diagram for explaining the model reconstruction step.
[0110] In the design knowledge extraction phase S3, the CPU 3 executes a model reconstruction step. Specifically, in this model reconstruction step, the performance factor X j The design variables x i Among them, the design variables x that do not appear in the response y as linear terms i The CPU 3 eliminates the nonlinear term consisting of
[0111] For example, as shown in the upper part of Fig. 12, five performance factors x1, x3, x5, x1·x3, and x2·x4 are extracted by the performance factor extraction phase S2, and one intercept β0 and contributions β1 to β5 to be multiplied by each performance factor are obtained. In this case, the first three of the five performance factors correspond to linear terms, i.e., main effects, and the remaining two correspond to nonlinear terms, particularly interaction terms. Here, x1 to x5 are all standardized.
[0112] In this case, when we look at the design variables x1 to x4, we see that there are five performance factors X j Of these, x1 and x3 appear in response y as both main effects and interaction terms, whereas x1 and x4 do not appear in response y as main effects, but only as an interaction term (f1 = x2 x4).
[0113] Now consider the case where only x1 and x3 are varied while x2, x4, and x5 are fixed. In this case, when a design variable such as x1 appears in both the main effect and the interaction term, the magnitude of the response y changes linearly overall due to the influence of the main effect, but as shown in the left diagram in Figure 12, the slope is thought to increase or decrease depending on the magnitude of x3. This can be interpreted as the influence of the interaction acting between x1 and x3. In this case, when an interaction term exists, the main effect is not determined, and it can be seen that the slope of the main effect changes depending on the value of the other design variable, x3, which forms the interaction.
[0114] On the other hand, if we consider the case where only x2 and x4 are varied while x1, x3, and x5 are fixed, when there is no main effect such as design variable x4 and only the interaction term f1 exists, the magnitude of the response y will depend entirely on the value of design variable x2 that appears in the interaction term f1.
[0115] For example, if the structure is considered to be a car body and the design variables are the plate thicknesses of the car body components, the above-mentioned trend can be interpreted from an engineering perspective as meaning that when it appears in both the main effect and interaction terms, such as design variable x1, the plate thickness of the part corresponding to design variable x1 affects the performance represented by response y, but the degree of influence on performance changes depending on the balance of the plate thickness with other parts corresponding to design variable x3.
[0116] On the other hand, when only an interaction term appears, such as design variable x4, it can be interpreted that the thickness of the part corresponding to design variable x1 itself does not affect performance, but rather affects performance only depending on the thickness of another part corresponding to design variable x2.
[0117] However, when considering vehicle body structure, it is difficult to imagine a structure that affects performance solely through its dependency on other parts, and even if such a structure did exist, its overall impact would likely be relatively low compared to one that has a main effect.
[0118] Therefore, in the model reconstruction step, performance factors that do not have a corresponding main effect, such as the interaction term f1 shown in Fig. 12, are deleted to reconstruct the model obtained by the above formulation. Although detailed illustration is omitted here, in this model reconstruction step, the CPU3 executes the multiple regression analysis again to update the contribution rate to be multiplied by each performance factor. Note that the timing for performing the multiple regression analysis may be after the performance factor extraction phase S2 is once completed, or may be during the performance factor extraction phase S2, as shown in Fig. 3.
[0119] In the latter case, a performance factor X based on the selection probability as illustrated in step S25 jWhen extracting the design factor X, which appears as a main effect, j After extracting the design factor X j (Design factor X appears as a main effect j ) consisting of design variables x i It may be configured to incorporate nonlinear terms including:
[0120] -Specific examples of the preliminary narrowing down phase S1 and the performance factor extraction phase S2- Fig. 9 is a diagram showing a specific example of the performance factor extraction phase S2. Fig. 10 is a diagram showing the influence of the interaction term on the crash performance value. The inventors of the present application verified the validity of the pre-narrowing phase S1 and the performance factor extraction phase S2 by numerical calculation. When verifying the validity, the true coefficient value (contribution degree) of the artificially generated nonlinear function and the contribution degree β calculated by the above two phases were j The artificially created real function has a total of 20 performance factors X j (j=1,2,...,20), which is composed of 15 linear terms (main effects), 4 interaction terms as nonlinear terms, and 1 other nonlinear term, and 50 samples are generated using this nonlinear function. Noise is added to the objective variable values of the sample data, and 4 dummy performance factors (dummy factors) that do not affect the output are added. The results of applying the pre-narrowing phase S1 and performance factor extraction phase S2 to the generated sample data correspond to the first column from the right in Figure 10. The second column from the right in Figure 10 shows the results of applying the conventional method, and the third column from the right corresponds to the true count values.
[0121] In addition, in the first column from the left, “X_i(i=1,2,…,20)” is the design variable x i In addition, each cell has a coefficient value (contribution rate β j ) is graphed. As shown in FIG. 10, the contribution rate β obtained when the method according to the present invention is applied is jThe values of are closer to the true count values compared to the conventional method. It can also be seen that similar values are obtained for the nonlinear terms that have interaction effects, except for X_8*X_8. For X_8*X_8, the contribution of the linear term X_8, β j It is believed that this is compensated for by calculating a larger value.
[0122] In particular, in the conventional method, the coefficient of determination R 2 Although the value of is about 96%, unlike the method of the present invention, the coefficient values of the nonlinear terms are all set to zero. In the conventional method, it is not possible to incorporate the interaction term into the response y.
[0123] As mentioned above, the interaction term is considered to show the dependency between parts. For example, consider the crashworthiness of the vehicle body (especially the crashworthiness in a full-wrap frontal collision) as the response y, while the design variable x i In the case of considering the plate thickness of the body components, as shown in FIG. 11, consider the cases where only the plate thickness of part A is increased, where only the plate thickness of part B is increased, and where the plate thicknesses of both parts A and B are increased.
[0124] Here, as shown in graph V1 in Fig. 11, if only the plate thickness of part A is increased, the collision performance value deteriorates, and if only the plate thickness of part B is increased, the collision performance value improves. Here, when the plate thicknesses of both parts A and B are increased, if there is no interaction term between the design variable corresponding to part A and the design variable corresponding to part B, the collision performance value becomes the sum of the amount of change when only the plate thickness of part A is increased and when only the plate thickness of part B is increased (see graph V3 in Fig. 11).
[0125] On the other hand, when the thicknesses of both parts A and B are increased, if an interaction term exists between the design variables corresponding to part A and part B, the crashworthiness value will differ from the sum of the changes in the case where only the thickness of part A is increased and the case where only the thickness of part B is increased, as shown in graph V4 of FIG. 11. The occurrence of such a difference means that an interaction term exists. In the example shown in FIG. 11, part A and part B are parts that can transmit loads to each other, such as a B-pillar and a side sill or a roof rail. The magnitude of the load transmission at that time, in other words, the strength of the connection between the parts, can be characterized by the magnitude of the coefficient of the interaction term. The effect of the balance of the thicknesses between the parts on the crashworthiness value varies depending on the magnitude of the coefficient of the interaction term. This indicates that the balance of the thicknesses can greatly vary the transmission path of the energy.
[0126] <Formulation of structural performance> As described with reference to FIG. 8, the design support method according to this embodiment reduces the performance factor X j By applying multiple extraction methods to the set of j This selection order is determined by the performance factor X j It can be said that this shows the importance of the following. When setting the selection order, by using multiple extraction methods, rather than using only a specific extraction method, the selection order can be set more appropriately. And, the performance factor X with a relatively high selection order is j Regarding the performance factor X, which has a relatively low selection rank, it is actively incorporated into the model formula y that describes the performance of the structure. j By passively incorporating into the model formula y, the performance of the structure can be faithfully described while suppressing the number of terms used in the formulation. This makes it possible to formulate the performance of the structure simply and quickly.
[0127] In addition, as shown in Fig. 5, the performance factor X j contains the design variable x iNonlinear terms formed by multiplying each other, especially the interaction term x i x k This allows the model to include interactions related to the connections between parts, which is useful when modeling dynamic phenomena such as the crash performance of a structure.
[0128] As illustrated in FIG. 4, the design support method according to the present embodiment further includes a third aspect, in which a performance factor X based on the selection order is calculated. j Prior to the selection of j Here, as shown in Fig. 6, the design variable x i By eliminating the coefficients, the search can be narrowed down more quickly. At this time, by setting the number of levels of the tree structure T1 in advance, the search can be narrowed down appropriately for the structure to be supported. This allows for more efficient formulation.
[0129] As explained with reference to FIG. 12, the design variable x i Among these, the design variable x does not appear in the model formula as a linear term, but appears in the model only as an interaction term. i This step is performed to eliminate the performance factor X that should be considered. j Since the number of can be reduced, the amount of calculation can be reduced.
[0130] As explained in Figure 8, the frequently extracted performance factor X j For example, a performance factor that is extracted by all extraction methods is set to a high selection priority, while a performance factor that is extracted less frequently, for example, a performance factor that is extracted by only one of the extraction methods, is set to a low selection priority. By setting in this way, the selection priority can be set more appropriately, and the performance of the structure can be formulated more faithfully.
[0131] In that case, each performance factor X j Coefficient β multiplied by jBy determining the extraction frequency based on the level of , the extraction frequency can be clearly determined, which is advantageous in formulating the performance of the structure simply and quickly. In this case, by using sparse regression analysis and stepwise regression analysis, the coefficient β j This allows the user to more clearly judge whether the level is high or low.
[0132] In addition, as described with reference to Fig. 9, a more robust formulation can be achieved by using k-fold cross validation for formulation. This is effective in achieving both a simple and fast formulation of the performance of the structure and robustness of the model formula.
[0133] Other Embodiments In the above embodiment, a computer having one CPU 3 is exemplified as an example of the computer 1, but the present disclosure is not limited to this example. The computer 1 includes parallel computers such as supercomputers and PC clusters in addition to personal computers. For example, only steps that require a calculation time, such as the pre-narrowing phase S1 and the performance factor extraction phase S2 in Fig. 3, may be executed by a parallel computer, and the remaining steps, such as the design knowledge extraction phase S3, may be executed by a personal computer.
[0134] That is, the "computing unit" in this disclosure may be configured by combining a computing unit in a specific computer with a computing unit in another computer. In that case, "computer 1" refers to a computing system consisting of multiple computers. The same applies to the storage unit. [Industrial Applicability]
[0135] As described above, the present disclosure is useful in supporting the design of structures consisting of multiple parts, such as automobile bodies, and has industrial applicability. [Explanation of symbols]
[0136] 1. Computer 3 CPU (arithmetic unit) 7 RAM (memory section) Ds Learning data Dv verification data Pl division pattern S1 Pre-narrowing phase (pre-narrowing step) S2 Performance factor extraction phase S21 Read step S23 Extraction step S24 Ranking step S25 Formulation step S27 Index calculation step S29 Verification Step S3 Design knowledge extraction phase (model reconstruction step) T1 Tree structure x i Design Variables X j performance factor
Claims
1. A design support method for formulating the performance of a structure composed of a plurality of parts using design variables in which the specifications of each part are quantified, by using a computer having a calculation unit that executes a program and a storage unit that reads data, comprising: a reading step in which the storage unit reads a set of performance factors having one or more of the design variables as arguments; an extraction step in which the calculation unit applies a plurality of types of extraction methods to the set of performance factors to extract performance factors that contribute to the performance of the structure for each of the extraction methods; a ranking setting step in which the calculation unit sets a selection ranking of the performance factors based on the extraction results obtained by the extraction methods; a formulation step in which the calculation unit selects the performance factors from the set of performance factors according to the selection order, and combines the selected performance factors to formulate the performance of the structure. A design support method comprising:
2. 2. The design support method according to claim 1, The performance factor includes a linear term consisting of the design variables themselves and a nonlinear term consisting of the multiplication of the design variables. A design support method comprising:
3. 3. The design support method according to claim 2, a pre-narrowing step in which the calculation unit sets a set of the performance factors prior to the extraction step; In the pre-narrowing step, a genetic operation is performed on a tree structure having the design variables as terminal nodes and multiplication operators as non-terminal nodes to set the set of performance factors; In the pre-narrowing step, the genetic operation is performed so that the number of layers of the tree structure is equal to or less than a predetermined maximum depth. A design support method comprising:
4. The design support method according to claim 2 or 3, The method includes a model reconstruction step in which the calculation unit excludes, from the performance of the structure, the nonlinear terms consisting of design variables that do not appear in the performance of the structure as the linear terms, among the design variables that constitute the performance factor. A design support method comprising:
5. 5. The design support method according to claim 1, The calculation unit determines an extraction frequency for each of the performance factors based on the extraction results obtained for each of the extraction methods, The calculation unit sets the selection order based on the extraction frequency. A design support method comprising:
6. 6. The design support method according to claim 5, The performance of the structure is formulated by a linear combination of the performance factors; the calculation unit executes a multiple regression analysis using the performance factors as explanatory variables and the performance of the structure as a response variable for each of the plurality of types of extraction methods; The calculation unit determines the extraction frequency based on the coefficients of the explanatory variables calculated by the multiple regression analysis. A design support method comprising:
7. 7. The design support method according to claim 6, The multiple types of extraction methods include regression analysis with sparsity. A design support method comprising:
8. The design support method according to claim 6 or 7, The multiple extraction methods include stepwise regression analysis. A design support method comprising:
9. A design support method according to any one of claims 6 to 8, an index calculation step of calculating an evaluation index that characterizes the accuracy of the formulation in the formulation step; a verification step of verifying the accuracy of the formulation in the formulation step based on the evaluation index calculated in the index calculation step; The storage unit reads a plurality of sample data each corresponding to a set of the performance factors; The calculation unit divides the plurality of sample data into training data and validation data according to a predetermined division pattern; the calculation unit executes the extraction step, the ranking step, and the formulation step based on the learning data, and executes the index calculation step based on the remaining verification data; the calculation unit executes the extraction step, the rank setting step, the formulation step, and the index calculation step repeatedly for each changed division pattern while changing the division pattern a plurality of times, and then executes the verification step; In the verification step, the calculation unit selects one of the performances of the structure formulated for each of the division patterns based on the evaluation index calculated for each of the division patterns. A design support method comprising:
10. A design support method according to any one of claims 1 to 9, the structure is a vehicle body, The design variables are the plate thicknesses of the plurality of components. A design support method comprising:
11. A design support program for formulating the performance of a structure composed of a plurality of parts using design variables in which the specifications of each part are quantified, by causing a computer having a calculation unit for executing the program and a storage unit for reading data to execute the program, The computer includes: a reading step in which the storage unit reads a set of performance factors having one or more of the design variables as arguments; an extraction step in which the calculation unit applies a plurality of types of extraction methods to the set of performance factors to extract performance factors that contribute to the performance of the structure for each of the extraction methods; a ranking setting step in which the calculation unit sets a selection ranking of the performance factors based on the extraction results obtained by the extraction methods; the calculation unit selects the performance factors from the set of performance factors according to the selection order, and combines the selected performance factors to formulate the performance of the structure. A design support program comprising:
12. The design support program according to claim 11 is stored. A computer-readable storage medium comprising:
Citation Information
Patent Citations
Design supporting method for structure
JP2020071725A
Method of identifying technical design solutions
US20180011949A1
Design workflow construction device, design workflow construction method, design system, design method, design workflow construction program, and computer readable recording medium on which design workflow construction program is recorded
WO2010095418A1
