Girder design device, girder design method, and girder design computer program

Abstract

To appropriately output candidates for plate elements used in manufacturing a girder for supporting a floor slab of a bridge.SOLUTION: A girder designing device creates a plurality of candidates representing combinations of plate elements used in manufacturing a girder for supporting a floor slab of a bridge by determining values of variables, creates a variation candidate having, as a variable, a value calculated using a value of a variable in a candidate other than a provisional optimum candidate having the most preferred value of an evaluation function representing a weight of the girder among the plurality of candidates, replaces the provisional optimum candidate with the variation candidate if a value of the evaluation function for a variation-improvement candidate is preferable to the value of the evaluation function for the provisional optimum candidate, repeats the variation and the replacement until variation candidates whose values of the evaluation function are within a predetermined convergence range from the value of the evaluation function for the provisional optimum candidate are created a predetermined number of times of convergence, and outputs the provisional optimum candidate or the variation candidate as the optimum candidate after the variation candidates whose values of the evaluation function are within the predetermined convergence range from the value of the evaluation function for the provisional optimum candidate are created the number of times of convergence.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present disclosure relates to a girder design device, a girder design method, and a computer program for girder design that output candidates for plate elements used in manufacturing a girder that supports an upper load. [Background technology]

[0002] In order to manufacture girders such as plate girders and box girders that support surcharge loads, designers specify specifications such as the size and material strength of the plate elements that make up the girders so that the manufactured girders satisfy certain constraints. The design of girders is required to satisfy a wide range of constraints such as the yield strength, ultimate strength, buckling strength, rigidity, and fatigue strength of the girders.

[0003] Patent Document 1 describes a framework structure optimization design system that optimizes design variable data of a framework structure using a genetic search method. In the framework structure optimization design system described in Patent Document 1, the cross sections of beams that constitute the framework are used as design variables. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2001-134628 A Summary of the Invention [Problem to be solved by the invention]

[0005] In the frame structure optimization design device described in Patent Document 1, the cross section of a beam is used as a design variable. The frame structure optimization design device described in Patent Document 1 searches for a beam cross section that optimizes an evaluation function value using a genetic search method. On the other hand, when variables expressed by numerical values ​​such as the length, width, and plate thickness of plate elements constituting a girder are used as design variables, there is a certain relationship between the variation of the design variable and the evaluation function value. Therefore, when designing a girder, it is preferable to optimize the design variable taking into consideration the relationship between the variation of the design variable and the evaluation function value.

[0006] Therefore, an object of the present disclosure is to provide a girder design device that appropriately outputs candidates for plate elements to be used in manufacturing a girder that supports an upper load. [Means for solving the problem]

[0007] The girder design device disclosed herein comprises a creation unit that creates multiple candidates representing combinations of plate elements used in manufacturing a girder that supports an upper load by determining values ​​of variables representing at least the length, width, and plate thickness of each plate element; a change unit that creates change candidates having variables whose values ​​are calculated using the values ​​of the variables of candidates other than the tentative optimal candidate that has the most preferable value of an evaluation function representing the weight of the girder among the multiple candidates; a replacement unit that replaces the tentative optimal candidate with the change candidate if the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the tentative optimal candidate; and an output unit that repeats the change and replacement until a change candidate whose evaluation function value is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate is created a predetermined number of convergence times, and outputs the tentative optimal candidate or the change candidate as the optimal candidate after a change candidate whose evaluation function value is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate has been created a predetermined number of convergence times.

[0008] In the girder design device according to the present disclosure, it is preferable that the creation unit determines variables so that each of the multiple candidates satisfies a specified constraint condition, and the replacement unit does not replace the provisional optimal candidate with the change candidate if the change candidate does not satisfy the constraint condition.

[0009] It is preferable that the girder design device according to the present disclosure further includes a discretization unit that discretizes variables in each of the multiple candidates.

[0010] In the girder design device according to the present disclosure, it is preferable that the creation unit defines the variables to include the number of main girders to which plate elements are connected in the longitudinal direction of the bridge and which are arranged in the width direction of the bridge.

[0011] The girder design method disclosed herein includes creating a plurality of candidates representing combinations of plate elements used in manufacturing a girder that supports an upper load by determining values ​​of variables representing at least the length, width, and plate thickness of each plate element, creating a change candidate having as variables values ​​calculated using the values ​​of the variables of a candidate other than a tentative optimum candidate that has the most preferable value of an evaluation function representing the weight of the girder among the plurality of candidates, replacing the tentative optimum candidate with the change candidate if the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the tentative optimum candidate, repeating the change and replacement until a change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimum candidate has been created a predetermined number of times, and after a change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimum candidate has been created a predetermined number of times, outputting the tentative optimum candidate or the change candidate as the optimum candidate.

[0012] The computer program for girder design disclosed herein causes a processor to execute the following steps: create multiple candidates representing combinations of plate elements used in manufacturing a girder that supports an upper load by determining values ​​of variables representing at least the length, width, and plate thickness of each plate element; create change candidates having variables whose values ​​are calculated using the values ​​of the variables of a candidate other than the tentative optimal candidate that has the most preferable value of an evaluation function representing the weight of the girder among the multiple candidates; replace the tentative optimal candidate with the change candidate if the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the tentative optimal candidate; repeat the change and replacement until a change candidate whose evaluation function value is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate has been created a predetermined number of times, and after a change candidate whose evaluation function value is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate has been created a predetermined number of times, output the tentative optimal candidate or the change candidate as the optimal candidate. Effect of the Invention

[0013] According to the girder design device of the present disclosure, it is possible to appropriately output candidates for plate elements to be used in manufacturing a girder that supports an upper load. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 is a schematic diagram showing a schematic configuration of a computer. [Diagram 2] FIG. 2 is a functional block diagram of a processor included in the computer. [Diagram 3] (a) is a schematic diagram showing the side of the girder supporting the bridge deck, (b) is a schematic diagram showing the top surface of the girder, and (c) is a schematic diagram showing the AA cross section of the girder. [Figure 4] 1 is a process flowchart of a girder design method. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, a detailed description will be given of a girder design device that appropriately outputs candidates for plate elements used in manufacturing a girder that supports a surcharge load, with reference to the drawings. However, it should be understood that the present invention is not limited to the drawings or the embodiments described below.

[0016] The girder design device of the present disclosure optimizes a combination of plate elements used in manufacturing a girder that supports a surcharge load using a differential evolution method. The girder design device creates multiple candidates representing combinations of plate elements by determining values ​​of variables representing at least the length, width, and plate thickness of each plate element. The girder design device identifies a tentative optimal candidate having the most favorable value of an evaluation function among the multiple candidates. The girder design device creates a change candidate using the variable values ​​of candidates other than the tentative optimal candidate, and if the change candidate is more favorable than the tentative optimal candidate, replaces the tentative optimal candidate with the change candidate. The girder design device repeats the change and replacement until convergence, and then outputs the tentative optimal candidate or the change candidate as the optimal candidate.

[0017] FIG. 1 is a schematic diagram showing a schematic configuration of a computer.

[0018] A computer 1, which is an example of a girder design device, includes an input / output interface 11, a memory 12, and a processor 13.

[0019] The input / output interface 11 has an interface circuit for receiving data to be processed by the computer 1 or outputting data processed by the computer 1. The input / output interface 11 includes, for example, a communication interface circuit for connecting the computer 1 to a communication network, or a peripheral device interface circuit for connecting the computer 1 to various peripheral devices such as a keyboard, a mouse, and a display. The input / output interface 11 receives operation input from a terminal device via the communication interface circuit, or from an input device such as a keyboard or a mouse via the peripheral device interface circuit. The input / output interface 11 also outputs display data for displaying the output optimum candidate to the terminal device via the communication interface circuit, or to a display via the peripheral device interface circuit.

[0020] The memory 12 includes, for example, at least one of a semiconductor memory, a magnetic disk device, and an optical disk device. The memory 12 stores computer programs, data, and the like used for processing by the processor 13. The computer programs include a driver program, an operating system program, and a computer program for girder design. The computer programs may be provided in a form recorded on a computer-readable portable recording medium such as a semiconductor memory, a magnetic recording medium, or an optical recording medium.

[0021] The processor 13 includes one or more processors and their peripheral circuits. The processor 13 is a processing circuit that comprehensively controls the overall operation of the computer 1, and is, for example, a CPU (Central Processing Unit). The processor 13 controls the operation of the input / output interface 11, etc., so that various processes of the computer 1 are executed by appropriate means based on computer programs, etc. stored in the memory 12. The processor 13 executes processes based on the various programs stored in the memory 12. The processor 13 can also execute a plurality of various programs in parallel.

[0022] 2 is a functional block diagram of the processor 13 included in the computer 1. The processor 13 includes a creation unit 131, a discretization unit 132, a modification unit 133, a replacement unit 134, and an output unit 135. Each of these units included in the processor 13 is a functional module implemented by a program executed on the processor included in the processor 13. Alternatively, each of these units included in the processor 13 may be implemented in the computer 1 as an independent integrated circuit, microprocessor, or firmware.

[0023] The creation unit 131 creates a plurality of candidates representing combinations of plate elements used in manufacturing a girder that supports a surcharge load. The creation unit 131 identifies the plate elements included in the combination by determining at least the values ​​of variables representing the size of each plate element. The plate elements are, for example, steel materials.

[0024] Figure 3(a) is a schematic diagram showing the side of the girder supporting the bridge deck, Figure 3(b) is a schematic diagram showing the top surface of the girder, and Figure 3(c) is a schematic diagram showing the AA cross section of the girder.

[0025] The girder G is installed between the abutments AB1 and AB2, and supports the load including the deck FS. The girder G is connected to the abutment AB1 via the support SH11, and to the abutment AB2 via the support SH12. The length of the girder G is L G The distance from the center of thickness of the web of the leftmost block B42 to the center of thickness of the web of the rightmost block B12 is W G It is.

[0026] The girder G has four main girders. Each main girder is made up of multiple blocks of one or more plate materials connected by joints. For example, B11 Block B11 and length L B12 Block B12 is connected to block B12 by joint J11, and block B12 is connected to block B12 by joint J11. B13 It is connected to block B13 by joint J12 and has a length of L G The main girders are constructed as follows.

[0027] The main girders of girder G are connected to each other by cross girders arranged in the width direction. For example, the main girder in which blocks B12 and B13 are connected in the longitudinal direction of the bridge by joint J12 is connected to the main girder in which blocks B22 and B23 are connected in the longitudinal direction of the bridge by joint J22 by cross girders CB14, CB15, CB16, and CB17. Similarly, the main girder in which blocks B22 and B23 are connected by joint J22 is connected to the main girder in which blocks B32 and B33 are connected by joint J32 by cross girders CB25, etc. Similarly, the main girder in which blocks B32 and B33 are connected by joint J32 is connected to the main girder in which blocks B42 and B43 are connected by joint J42 by cross girders CB35, etc. Each cross girder has a length d in the longitudinal direction of the bridge. L The electrodes are spaced apart from each other.

[0028] The plate thickness of each of the n plate elements constituting the block B42 is T B421 , T B422 , …, T B42n The plate thickness of each of the plate elements constituting the block B43 connected to the block B42 may be the same as or different from the plate thickness of each of the plate elements constituting the block B42.

[0029] The web width of block B12 is W B12W The width of the webs of blocks B22, B32, B42, which are arranged parallel to block B12 along the longitudinal direction of the bridge, is W B12W It may be the same as or different from.

[0030] The creation unit 131 creates a plurality of candidates representing combinations of plate elements used in the manufacture of the girder G by {(L B11 , W B11W , T B111 ), (L B11 , W B11W , T B112The creation unit 131 may determine each variable by a random number. The creation unit 131 may determine each variable by using a random number set to generate a numerical value within a predetermined range.

[0031] The creation unit 131 determines variables so that a candidate for a combination of plate elements that satisfies a predetermined constraint is selected. The constraint can be, for example, the length L of the girder G, which is calculated as the total length of the plate elements connected in the longitudinal direction of the bridge. G The constraint conditions include that the distance between abutments AB1 and AB2 corresponds to the distance between abutments AB1 and AB2, and that the strength of the girder G manufactured from the plate elements included in the combination is greater than a predetermined strength threshold. The constraint conditions may be existing design standards such as the Highway Bridge Specifications, AASHTO LFRD Bridge Design Specification, and Euro Code. The creation unit 131 inputs variables included in the candidates into a known algorithm to calculate a value indicating the standards that the girder G should satisfy, including the strength of the girder G. The creation unit 131 may discard candidates that do not satisfy the constraint conditions, and repeat the creation of new candidates and the determination of whether or not they satisfy the constraint conditions until a desired number of candidates are obtained. The creation unit 131 may also impose a penalty on the evaluation value of a candidate that does not satisfy the constraint conditions so that the candidate is less likely to be selected.

[0032] The creation unit 131 determines the variables so as to include the number of main girders arranged in the width direction of the bridge, to which the plate elements are connected in the longitudinal direction of the bridge. In the example of Fig. 3, girder G has four main girders arranged in the width direction of the bridge.

[0033] The discretization unit 132 discretizes the variables in each of the multiple candidates. The discretization unit 132 changes the value of the variable in each of the multiple candidates to the closest discrete value among the multiple discrete values ​​stored for each variable in the memory 12. For example, if 9 mm, 12 mm, and 15 mm are stored in the memory 12 as multiple discrete values ​​for plate thickness, and a variable corresponding to the plate thickness of a certain plate element in a certain candidate is 11.2 mm, the discretization unit 132 changes the value of the variable to 12 mm, which is the closest to 11.2 mm.

[0034] The varying unit 133 identifies a provisional optimum candidate having the most preferable value of an evaluation function representing the weight of a digit from among the multiple candidates.

[0035] The modification unit 133 obtains the volume of each plate element by multiplying the length, width, and plate thickness of each plate element among the variables included in each of the multiple candidates, and obtains the sum of the volumes of the plate elements. The modification unit 133 then multiplies the sum of the volumes of the plate elements by the specific gravity of the plate elements stored in the memory 12 to obtain the total weight of the plate elements used in manufacturing the girder G in the candidate.

[0036] The change unit 133 obtains the number of joints used in one main girder by subtracting 1 from the number of plate elements that make up one main girder. The change unit 133 obtains the number of joints used in one main girder by multiplying the number of main girders arranged in the width direction of the bridge. The change unit 133 then obtains the total weight of joints used in manufacturing the girder G in that candidate by multiplying the number of joints used in the candidate by the weight of the joints stored in the memory 12.

[0037] The change unit 133 calculates the weight of the girder G, which is the sum of the total weight of the plate elements and the total weight of the joints, for each of the multiple candidates, and identifies the candidate with the most preferable weight of the girder G as the provisional optimum candidate. For example, when a girder is designed to reduce the load on the support, the change unit 133 identifies the candidate with the smallest value of the evaluation function representing the weight of the girder G as the most preferable provisional optimum candidate. Also, when a girder is designed to reduce the cost required for production, the change unit 133 identifies the candidate with the smallest value of the evaluation function representing the production cost as the most preferable provisional optimum candidate.

[0038] The change unit 133 creates a change candidate having, as a variable, a value calculated using the variable value of a candidate other than the provisional optimum candidate among the multiple candidates.

[0039] For example, the variation unit 133 selects three candidates other than the provisional optimal candidate from among the multiple candidates. Let the provisional optimal candidate x and the variables included in x be x=(x1, x2, ..., x n ), and the three candidates a, b, c and the variables included in each candidate are represented as a = (a1, a2, ..., a n ), b=(b1,b2,...,b n ), c=(c1,c2,...,c n ) is expressed as

[0040] In this case, the candidate changes y and the variables included in y are expressed as y=(y1,y2,...,y n ), the change unit 133 changes each variable y i With probability C, y i =x i Then, with probability 1-C, y i =a i +F(b i -c i ), where F is a constant related to the convergence speed of optimization, and C is a constant related to the substitution rate (crossover rate) of variables. The variation unit 133 creates variation candidates y using the values ​​of F and C stored in the memory 12 according to the settings of the user or administrator of the computer 1.

[0041] The replacement unit 134 calculates values ​​of the evaluation function for the provisional optimal candidate and the change candidate, and replaces the provisional optimal candidate with the change candidate if the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the provisional optimal candidate.

[0042] For example, when the weight of a digit of the change candidate is lighter than the weight of a digit of the temporary optimum candidate, the replacing unit 134 replaces the temporary optimum candidate with the change candidate.

[0043] The output unit 135 causes the change unit 133 and the replacement unit 134 to create change candidates and replace the tentative optimal candidates until all of the multiple candidates are used to create change candidates. In this disclosure, the period from when a change candidate is first created using a candidate selected from the multiple candidates to when all of the multiple candidates are used to create change candidates is referred to as a generation.

[0044] The output unit 135 determines the number of times in a generation in which a change candidate has been created in which the value of the evaluation function falls within a predetermined convergence range from the value of the evaluation function for the provisional optimal candidate and stored in the memory 12. The output unit 135 determines whether the number of times a change candidate has been created in which the value of the evaluation function falls within the convergence range is equal to or greater than the predetermined convergence count stored in the memory 12.

[0045] If the number of times that change candidates for which the evaluation function value falls within the convergence range have been created is not equal to or greater than the convergence count, the output unit 135 causes the change unit 133 and the replacement unit 134 to create change candidates and replace the provisional optimal candidate.

[0046] When the number of times that the change candidates for which the evaluation function value falls within the convergence range are created is equal to or greater than the convergence count, the output unit 135 outputs the tentative optimum candidate or the change candidate as the optimum candidate. The output unit 135 outputs display data for displaying the optimum candidate to a display connected via the input / output interface 11, for example. The output unit 135 may also output display data for displaying the optimum candidate to a terminal device connected to a communication network connected via the input / output interface 11.

[0047] The output unit 135 may obtain the total number of times in all generations up to now in which change candidates have been created in which the value of the evaluation function falls within a predetermined convergence range from the value of the evaluation function for the provisional optimal candidate and which is stored in the memory 12. In this case, the memory 12 stores a value of the convergence count set for determining convergence based on the total number of times.

[0048] FIG. 4 is a process flow chart of the girder design method.

[0049] The processor 13 of the computer 1 executes the process shown in FIG. 4 every time it receives a user operation via the input / output interface 11 requesting optimization of the girder design.

[0050] First, the creation unit 131 of the processor 13 creates multiple candidates representing combinations of plate elements used in manufacturing girders supporting a bridge deck by determining values ​​of variables representing at least the length, width, and plate thickness of each plate element (step S1).

[0051] Next, the discretization unit 132 of the processor 13 discretizes the variables in each of the multiple candidates (step S2).

[0052] Next, the change unit 133 of the processor 13 identifies a provisional optimum candidate having the most preferable value of the evaluation function expressing the weight of the digit G from among the multiple candidates (step S3).

[0053] Furthermore, the change unit 133 creates a change candidate having, as a variable, a value calculated using the variable value of a candidate other than the provisional optimum candidate (step S4).

[0054] The replacement unit 134 of the processor 13 determines whether or not the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the provisional optimum candidate (step S5).

[0055] If it is determined that the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the provisional optimal candidate (step S5: Y), the replacement unit 134 replaces the provisional optimal candidate with the change candidate (step S6), and the processing of the processor 13 proceeds to step S7 described below.

[0056] When it is not determined that the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the provisional optimum candidate (step S5: N), the processor 13 proceeds to the process of step S7 described later.

[0057] The output unit 135 of the processor 13 determines whether or not a change candidate, in which the value of the evaluation function falls within a predetermined convergence range from the value of the evaluation function for the provisional optimum candidate, has been created a predetermined number of times (step S7).

[0058] When it is determined that the change candidates falling within the predetermined convergence range have not been created a convergence number of times (step S7: N), the process of the processor 13 returns to step S2.

[0059] If it is determined that a change candidate that falls within the predetermined convergence range has been created a convergence number of times (step S7: Y), the output unit 135 outputs the tentative optimum candidate or the change candidate as the optimum candidate, and ends the girder design process.

[0060] An example of the computer 1 of the present disclosure is shown below. In this example, the length L G The design was carried out for three different digits with different conditions for each case.

[0061] In cases 1-5, the length L of digit G G The length of the girder is 25m, the number of girders is 4, and the spacing between the girders is 2.44m. B11 = 6.9m, L B12 = 11.2m, L B13 =6.9m and divided into three blocks.

[0062] In Case 6-10, the length L of digit G GThe length of the girder is 35m, the number of girders is 4, and the spacing between the girders is 2.44m. B11 = 9.7m, L B12 = 15.6m, L B13 = 9.7m and divided into three blocks.

[0063] In cases 11-15, the length L of digit G G The length of the girder is 45m, the number of girders is 4, and the spacing between the girders is 2.44m. B11 = 12.5m, L B12 = 20.0m, L B13 =12.5m and divided into three blocks.

[0064] Length L G For each of the three cases with different digits, the design was carried out under one of the following three conditions.

[0065] Condition 1 is digit height W B The upper and lower flange widths were fixed, and the web thickness was reduced by 1 mm each time to find the minimum thickness that passed all inspections, and the lower flange thickness and upper flange thickness were determined in the same manner. Condition 1 assumes the conventional design procedure.

[0066] Conditions 2 and 3 were designed using the differential evolution method with the computer 1 of the present disclosure. The optimization conditions using the differential evolution method were F = 0.8, the number of individuals per generation = 30, and the termination condition = σ / μ≦10 -7 Or the number of generations was set to 10,000 or more. Here, σ and μ are the standard deviation and average, respectively, of the evaluation function values ​​calculated from all individuals in the generation. The evaluation function was set to the value obtained by multiplying the steel weight by the penalty from the design review, and the minimization problem was solved.

[0067] In condition 2, each design variable was a real number.

[0068] In condition 3, the plate thickness, which is one of the design variables, was set in inch pitches (market size pitches). Condition 3 corresponds to discretization of variables by the discretization unit 132. Note that the inch pitches are pitches in mm that are 9, 12, 16, 19, 22, 25, 28, 32, 36, 38, 40, and 45, and for 45 mm or more, the pitch is 5 mm.

[0069] For comparison with the results of condition 3, the plate thicknesses of the design results for conditions 1 and 2 were also discretized by rounding up to inch pitches, making them different cases from the cases in which such adjustments were not made.

[0070] The design cases are listed in Table 1 below.

[0071] [Table 1]

[0072] The results for each case are shown in Table 2 below.

[0073] [Table 2]

[0074] As shown in Table 2, regardless of the length of the girder to be designed, the design results under conditions 2 and 3 using the computer 1 of the present disclosure were able to reduce the steel weight compared to the design results under condition 1. The weight reduction rate of the cases where the plate thickness was not discretized under condition 2 (cases 3, 8, and 13) was 4% to 9% compared to the cases where the plate thickness was not discretized under condition 1 (cases 1, 6, and 11). The weight reduction rate of the cases where the plate thickness was discretized under condition 3 (cases 5, 10, and 15) was 9% to 14% compared to the cases where the plate thickness was discretized under condition 1 (cases 2, 7, and 12). In addition, the weight reduction rate of the cases where the plate thickness was discretized under condition 3 (cases 5, 10, and 15) was 1% to 9% compared to the cases where the plate thickness was discretized under condition 2 (cases 4, 9, and 14).

[0075] In this way, the computer 1 was able to appropriately output candidates for plate elements to be used in manufacturing girders that support the surcharge load.

[0076] It should be understood that those skilled in the art can make various changes, substitutions, and alterations thereto without departing from the spirit and scope of the present disclosure. [Explanation of symbols]

[0077] 1. Computer 13 Processors 131 Creation Department 132 Discretization part 133 Changes 134 Substitution part 135 Output section

Claims

1. A creation unit that creates a plurality of candidates representing combinations of plate elements used in manufacturing a girder that supports a surcharge load by determining values ​​of variables that represent at least the length, width, and plate thickness of each plate element; a change unit that creates change candidates having variables that are calculated using values ​​of the variables in candidates other than the tentative optimum candidate that has the most preferable value of an evaluation function that represents the weight of the digit among the plurality of candidates; a replacement unit that replaces the provisional optimal candidate with the change candidate when a value of the evaluation function for the change candidate is more preferable than a value of the evaluation function for the provisional optimal candidate; an output unit that repeats the creation and replacement of the change candidate until the change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate is created a predetermined number of times, and outputs the tentative optimal candidate or the change candidate as an optimal candidate after the change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate is created a predetermined number of times; A girder design device comprising:

2. The creation unit determines the variables such that each of the plurality of candidates satisfies a predetermined constraint condition; The girder design device according to claim 1 , wherein the replacement unit is configured not to replace the tentative optimum candidate with the change candidate if the change candidate does not satisfy the constraint condition.

3. The girder design device according to claim 1 , further comprising a discretization unit that discretizes variables in each of the plurality of candidates.

4. The girder design device according to any one of claims 1 to 3, wherein the creation unit determines the variables to include the number of main girders to which the plate elements are connected in the longitudinal direction of the girder and which are arranged in the width direction of the girder.

5. Creating a plurality of candidates representing combinations of plate elements to be used in the manufacture of a girder supporting a surcharge load by determining values ​​of variables representing at least the length, width, and plate thickness of each of the plate elements; Create a change candidate having a variable that is calculated using the value of the variable in a candidate other than the provisional optimum candidate that has the most preferable value of the evaluation function representing the weight of the digit among the plurality of candidates; if the value of the evaluation function for the change candidate is more preferable than the value of the evaluation function for the interim optimal candidate, replacing the interim optimal candidate with the change candidate; repeating the creation and replacement of the change candidates until the change candidates for which the value of the evaluation function falls within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate stored in a computer memory have been created a predetermined number of times, and outputting display data for displaying the tentative optimal candidate or the change candidates as optimal candidates after the change candidates for which the value of the evaluation function falls within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate have been created a predetermined number of times. The girder design method further comprises causing a computer to execute the steps of:

6. Creating a plurality of candidates representing combinations of plate elements to be used in the manufacture of a girder supporting a surcharge load by determining values ​​of variables representing at least the length, width, and thickness of each plate element; creating a change candidate having a variable that is calculated using a value of the variable in a candidate other than the tentative optimum candidate having the most preferable value of the evaluation function representing the weight of the digit among the plurality of candidates; replacing the interim optimal candidate with the change candidate if a value of the evaluation function for the change candidate is more preferable than a value of the evaluation function for the interim optimal candidate; repeating the creation and replacement of the change candidate until the change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate is created a predetermined number of times, and after the change candidate whose value of the evaluation function is within a predetermined convergence range from the value of the evaluation function for the tentative optimal candidate is created a predetermined number of times, outputting the tentative optimal candidate or the change candidate as an optimal candidate; A computer program for girder design that causes a processor to execute the above.

Images