JP2014153995A - Design method of architectural structure, manufacturing method using the same, and design device using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a design method of a frame body capable of determining the optimal solution of a design factor in a short time.SOLUTION: The design method is a method for designing, using a computer, a frame body 2 of an architectural structure including pillars 3 and beams 4. The beams 4 include a plurality of large beams 7 horizontally interconnecting the pillars 3, and small beams 8 for further dividing a horizontal frame surface 9 surrounded by the large beams 7. This design method includes an optimizing calculation process of calculating at least one optimal solution of a design factor for at least one small beam 8 on the basis of a genetic algorithm.

Description

  The present invention provides a frame design method capable of obtaining an optimal solution of a design factor in a short time.

  For example, a structure of a building by a shaft construction method has structural members including columns, beams, bearing walls, and the like. Further, the beam includes a plurality of large beams that horizontally connect between the columns, and a small beam that further classifies the horizontal frame surrounded by the large beams. Design factors including the arrangement and cross-sectional shape of the beam are determined according to the shape of the building and the load conditions. 2. Description of the Related Art Conventionally, a design factor that can increase the strength of a frame body while satisfying the shape and load conditions of a building has been obtained using a computer. Related technologies include the following.

Japanese Patent No. 4712075

  In order to obtain an optimal solution of the design factor, for example, it is conceivable to calculate the strength of the frame for a plurality of samples in which all patterns of the design factor are combined. However, this method has a problem in that a lot of calculation time is required because calculation or determination processing is required for all samples.

  On the other hand, in the method described in Patent Document 1, a method of artificially reducing the number of samples in order to shorten the calculation time has been proposed. However, in such a method, it is very difficult to narrow down the sample appropriately, and there is a problem that it is difficult to obtain an optimal solution of the design factor in a short time.

  The present invention has been devised in view of the actual situation as described above, and a design method for a frame body capable of obtaining an optimal solution of a design factor in a short time, a manufacturing method using the same, and a design apparatus used therefor The main purpose is to provide

  The invention according to claim 1 of the present invention includes a column and a beam, and the beam further includes a plurality of large beams that horizontally connect the columns and a small horizontal frame that is surrounded by the large beams. A method for designing a structure of a building including a beam using a computer, wherein the computer uses at least one optimum design factor based on a genetic algorithm for the at least one beam. It includes an optimization calculation step for calculating a solution.

  The invention according to claim 2 is the frame structure designing method according to claim 1, wherein the design factor includes an arrangement or a cross-sectional shape of the beam.

  According to a third aspect of the present invention, the optimal solution is the design factor that satisfies both the first target variable and the second target variable independent of each other that can be calculated from the frame. This is a design method for the frame.

  According to a fourth aspect of the present invention, there is provided the step of inputting basic building information including the shape and load condition of the building to the computer, and the computer to dispose the small beam on the frame. The design calculation condition, and the computer generates a group composed of a plurality of types of chromosome information with different design factors based on the design constraint condition, and the optimization calculation process The method for designing a frame according to claim 1, wherein the optimum solution is calculated using the group.

  In the invention according to claim 5, the optimization calculation step is a calculation step of calculating a first target variable and a second target variable of each frame based on the chromosome information, and at least a part of the calculation step. A gene manipulation step of crossing and mutating chromosome information, and a step of determining whether or not there is at least one chromosome information satisfying both the first target variable and the second target variable. 4 is a method of designing a frame body according to item 4.

  In the invention according to claim 6, the group includes an elite group including chromosome information having a high degree of optimization of the first target variable and the second target variable. In the genetic manipulation step, the elite group includes 6. The frame structure designing method according to claim 5, wherein the chromosome information is included in the new group without being crossed and mutated.

  The chromosome information may include at least one genetic locus capable of storing genetic information defining an arrangement of one beam for each horizontal frame, and the genetic information Includes direction information that defines the longitudinal direction of the beam, and position information that defines a fixed position of the beam, and the gene locus is a direction gene locus that can store the direction information, and the position The position information locus which can store information, and the chromosome information has the direction locus and the position locus for each of the small beams of each horizontal frame. This is a method for designing a frame.

  In the invention according to claim 8, in the design constraint condition, an orientation and a position where the beam can be fixed are set, and in the group generation step, the computer determines the design constraint condition based on the design constraint condition. The frame structure design method according to claim 7, further comprising a step of storing the direction information and the position information in the direction locus and the position locus.

  In the invention according to claim 9, the design constraint condition is set to an upper limit value of the number of the small beams arranged on each horizontal frame surface, and the upper limit value can be arranged on each horizontal frame surface The chromosomal information is defined by the same number of the directional loci and the positional loci as the upper limit value in each horizontal frame plane. 8. The method for designing a frame according to 8.

  In the invention according to claim 10, each horizontal frame is divided into a plurality of small frames by the small beams, and the small beams are arranged on the horizontal frame or the small frames. The direction information of each beam is composed of letters or numerical values having the same number of digits as the number of horizontal frames or beam frames on which the beam can be arranged, and each digit of the direction information is stored in each digit. The structure design method according to claim 7, wherein presence or absence of the small beam is defined on the horizontal frame surface or the small frame surface by the character or the numerical value.

  Further, in the invention described in claim 11, each of the horizontal frame surfaces is divided into a plurality of small frame surfaces by the small beams, and the small beams are arranged on the horizontal frame surface or the small frame surface, The gene information further includes a frame surface information for specifying the horizontal frame surface or the frame surface on which the beam is arranged, and the gene locus further includes a frame surface locus capable of storing the frame surface information. A frame structure designing method according to any one of claims 7 to 9.

  The invention according to claim 12 is characterized in that the optimization calculation step calculates a first target variable and a second target variable of each frame based on each chromosome information, and at least a part of the calculation step. A gene manipulation step of crossing and mutating the chromosome information, wherein the genetic information calculates, and crosses and mutates, the first target variable and the second target variable in each of the horizontal frames. Including the number information defining the number of the beam, wherein the locus further includes a number locus capable of storing the number information, and each chromosome information is defined as one number locus, and the calculation The process and the gene manipulation process are the first target variable and the second target only for the gene loci of the traversed beams corresponding to the number set in the number information in each horizontal frame plane. Calculation of characteristic variables, as well as a method for designing a rack structure according to any one of claims 7 to 11 to crossover and mutation.

  According to a thirteenth aspect of the present invention, the frame and the building are manufactured based on at least one of the optimum solutions calculated by the frame structure designing method according to any of the first to twelfth aspects. It is a manufacturing method of a building.

  A fourteenth aspect of the present invention is a design apparatus including an arithmetic processing unit for executing the frame design method according to any one of the first to twelfth aspects.

  The frame structure design method of the present invention is a method for designing a building frame using a computer. The building frame includes columns and beams. The beam includes a plurality of large beams that horizontally connect the columns, and a small beam that further divides the horizontal frame surrounded by the large beams. In the design method of the present invention, the computer includes an optimization calculation step of calculating at least one optimal solution of the design factor for at least one beam based on a genetic algorithm.

  In the design method of the present invention, the design factor can be evolved with a small number of samples without calculating a plurality of samples combining all patterns of the design factor. Therefore, the design method of the present invention can obtain the optimum solution of the design factor in a short time.

It is a perspective view of the design apparatus which performs the design method of this embodiment. It is a perspective view of a frame. It is a top view of a frame. (A) is a top view of the frame on which the first beam is arranged, (b) is a plan view of the frame on which the second beam is arranged. It is a flowchart which shows an example of the process sequence of the design method of this embodiment. It is a flowchart which shows an example of the process sequence of the basic information input process of this embodiment. It is a perspective view of a frame without a small beam. It is a top view which shows the node of a horizontal frame. It is a conceptual diagram of the group which consists of several chromosome information. It is a flowchart which shows an example of the process sequence of the group production | generation process of this embodiment. (A), (b) is a diagram which shows the direction information and position information of a 1st beam, and the 1st beam arranged on a horizontal frame. (A), (b) is a diagram which shows the direction information and position information of a 2nd small beam, and the 2nd small beam arrange | positioned on a horizontal frame. It is a flowchart which shows an example of the process sequence of the optimization calculation process of this embodiment. It is a diagram which shows the calculation result of a 1st target variable and a 2nd target variable. It is a flowchart which shows an example of the process sequence of the gene operation process of this embodiment. It is a flowchart which shows an example of the process sequence of the next generation group production | generation process of this embodiment. (A) is a conceptual diagram showing chromosome information before crossover, (b) is a conceptual diagram showing chromosome information after crossover. It is a conceptual diagram explaining a mutation. It is a conceptual diagram which shows the chromosome information of other embodiment. It is a diagram which shows the frame surface information of a 1st small beam, and the 1st small beam arrange | positioned on a horizontal frame surface. (A), (b) is a diagram which shows the frame surface information of a 2nd small beam, and the 2nd small beam arrange | positioned on a horizontal frame surface. (A), (b) It is a conceptual diagram which shows the chromosome information of other embodiment.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
The frame structure design method of the present invention (hereinafter sometimes simply referred to as “design method”) is a method for designing a frame structure of a building such as an industrialized house using a computer.

  As shown in FIG. 1, the computer 1 includes a main body 1a, a keyboard 1b, a mouse 1c, and a display device 1d. The main body 1a is provided with, for example, an arithmetic processing unit (CPU), a ROM, a working memory, a storage device such as a magnetic disk, and disk drive devices 1a1, 1a2.

  The storage device stores in advance processing procedures (programs) of the design method of the present embodiment. This processing procedure is executed by the arithmetic processing unit of the computer 1. Therefore, the computer 1 is configured as a design apparatus 1A for implementing the design method of the present invention.

  As shown in FIGS. 2 and 3, the frame body 2 includes a structural member 6 including, for example, a column 3, a beam 4, and a load bearing wall 5. The beam 4 includes a plurality of large beams 7 that horizontally connect the columns 3 and 3, and a small beam 8 that further divides the horizontal frame 9 surrounded by the large beams 7. The large beam 7 and the small beam 8 are configured as a floor beam for supporting a floor (not shown) of two or more floors or a roof beam for supporting the roof 10. The large beam 7 and the small beam 8 are made of, for example, a substantially horizontal H-shaped steel.

  As shown in FIG. 4A, each horizontal frame 9 is provided with two small beams by arranging a first beam (hereinafter also referred to as “first beam”) 8a. It is divided into a frame surface 13. Further, as shown in FIG. 4B, the horizontal frame 9 has a second beam (hereinafter also referred to as “second beam”) 8 b of the two beams 13. By being arranged in any one, it is divided into three small frames 13. Thus, the horizontal frame 9 is divided into a plurality of small frames 13 by arranging the small beams 8.

  In the figure, the small frame 13 of the present embodiment has a first small frame 13A closest to the reference point 15a which is the lower left apex of the horizontal frame 9, and a reference point 15a next to the first small frame 13A. The second small frame 13B is included. Further, the small structure surface 13 includes a third small structure surface 13C that is next to the reference point 15a after the second small structure surface 13B.

  FIG. 5 shows a specific processing procedure of the design method of the present embodiment. In this design method, at least one optimum solution of the design factor is calculated for at least one beam 8 (shown in FIG. 2). The design factor of this embodiment is the arrangement of the small beams 8.

  In the design method of this embodiment, first, basic information of a building is input to the computer 1 (basic information input step S1). FIG. 6 shows a specific processing procedure of the basic information input step S1 of the present embodiment.

  In the basic information input step S1 of the present embodiment, first, the shape of the building is input to the computer 1 (step S11). In this step S11, as shown in FIG. 7, for example, the shape and arrangement of the building pillar 3, the bearing wall 5, the girder 7, the horizontal frame 9 and the roof 10 (shown in FIG. 3) are transferred to the computer 1. Entered. In this step S11, the beam 8 (shown in FIGS. 2 and 3), which is a design factor, is not input.

  The columns 3, the load-bearing walls 5, and the girder 7 are set in terms of arrangement, length, and the like with reference to a predetermined horizontal module or vertical module. Further, the pillar 3 and the large beam 7 are fixed by, for example, a pin 16 in which a bolt is modeled. Thereby, the frame 2 and the horizontal frame 9 excluding the small beam 8 (shown in FIG. 2) are set. Furthermore, the long beam 7 of the present embodiment is arranged such that its longitudinal direction is along the X-axis direction or the Y-axis direction.

  In addition, parameters necessary for the structural calculation such as the cross-sectional shape and the secondary moment of the cross-section are set for the column 3, the bearing wall 5, and the large beam 7. The arrangement and parameters of the columns 3, the beams 7 and the bearing walls 5 are all stored in the computer 1 as numerical data.

  The horizontal frame 9 of the present embodiment includes a first horizontal frame 9A to a fourth horizontal frame 9D that supports the floor on the second floor, and a fifth horizontal frame 9E to the fourth horizontal frame 9E that supports the roof 10 (shown in FIG. 3). 8 horizontal frame 9H. Furthermore, as shown in FIG. 8, a plurality of nodes 17 arranged at equal intervals according to the horizontal module are defined on each horizontal frame 9A to 9H. The node 17 of the present embodiment is set only in a region excluding an opening (not shown) such as a staircase or a stairwell. Such horizontal frame surfaces 9 </ b> A to 9 </ b> H are stored in the computer 1.

  The arrangement and shape of the roof 10 (shown in FIG. 3) are set on the basis of the horizontal module or the vertical module, similarly to the pillar 3 and the beam 4. Such an arrangement of the roof 10 is stored in the computer 1 as numerical data.

  Next, in this embodiment, building load conditions are input to the computer 1 (step S12). The load condition is information regarding external force acting on the building. The load conditions are input based on, for example, various building specifications such as exterior wall specifications, floor specifications, roofing materials, fire resistance specifications, earthquake resistance grades or wind resistance grades. Such a load condition is also numerical data and is stored in the computer 1.

  The basic building information can be set using software such as general CAD and a consistent structural calculation system. In the present embodiment, a two-story building is shown as an example. However, for example, a one-story building or a three-story or more building may be used.

  Next, design constraint conditions for placing the beam 8 on the frame 2 are input to the computer 1 (step S2). As design constraint conditions of the present embodiment, as shown in FIG. 8, directions and positions where the small beams 8 can be fixed are set on the horizontal frame surfaces 9 </ b> A to 9 </ b> H.

  The longitudinal direction of the small beam 8 of the present embodiment is set. This longitudinal direction is limited to the X-axis direction or the Y-axis direction in the drawing. The longitudinal direction is not limited to such a mode, and may be limited to any direction including, for example, a component in the X-axis direction and a component in the Y-axis direction.

  Further, the small beam 8 of the present embodiment is the end closest to the reference point 15a of the horizontal frame surface 9 (hereinafter, simply referred to as “reference end”) among the pair of ends 8t, 8t of the small beam 8. There is a fixed position of 18). The fixed position of the reference end 18 is limited to the nodes 17 defined on the horizontal frame surfaces 9A to 9H. As described above, the node 17 is set only in a region excluding an opening (not shown) such as a staircase. Thereby, the small beam 8 is arrange | positioned according to a horizontal module in the area | region except the opening part (illustration omitted) of each horizontal frame surface 9. As shown in FIG.

  Furthermore, in this embodiment, the upper limit value of the number of the small beams 8 arrange | positioned on each horizontal frame 9A-9H is set as a design constraint condition. This upper limit value is set to be smaller than the maximum number of small beams 8 that can be arranged on each of the horizontal frame surfaces 9A to 9H. Moreover, an upper limit can be arbitrarily set according to various conditions, such as intensity | strength required for the frame 2, and cost. In the present embodiment, for example, “3” is set. These design constraint conditions are stored in the computer 1 as numerical data such as coordinate values.

  Next, the computer 1 generates a group composed of a plurality of types of chromosome information having different design factors based on the design constraint condition (group generation step S3). As shown in FIG. 9, each chromosome information 21 defines at least one gene locus 22 capable of storing gene information that defines the arrangement of one beam 8 for each of the horizontal frames 9A to 9H. . FIG. 10 shows a specific processing procedure of the group generation step S3 of the present embodiment.

  In the group generation step S3 of the present embodiment, first, a small beam 8 (shown in FIG. 4) is defined in the computer 1 (step S31). The small beam 8 is numerical data in which a cross-sectional shape, a cross-sectional secondary moment, and the like are defined. Such numerical data is stored in the computer 1.

  Next, the locus 22 of the chromosome information 21 is set in the computer 1 (step S32). As shown in FIG. 9, the gene locus 22 of the present embodiment includes a direction gene locus 22 a that can store direction information and a position gene locus 22 b that can store position information.

  In addition, the chromosome information 21 is defined by the same number of directional loci 22a and position loci 22b as the upper limit value (for example, 3) of the small beams 8 in each of the horizontal frames 9A to 9H. Therefore, each chromosome information 21 is a directional locus for each beam 8 (in this embodiment, the first beam 8a, the second beam 8b, and the third beam 8c) on each horizontal frame 9A to 9H. 22a and position locus 22b are assigned.

  Next, the computer 1 stores gene information at each locus 22 of the chromosome information 21 (step S33). Direction information is stored in the direction locus 22a of the present embodiment. Furthermore, position information is stored in the position locus 22b.

The direction information defines the longitudinal direction of the beam 8. In addition, the direction information is defined by at least one digit character or numerical value, in this embodiment, at least one digit numerical value. Each numerical value that can be taken by the direction information of the present embodiment is 0 to 2, and details are as follows.
0: No beam arrangement 1: Parallel to X-axis direction 2: Parallel to Y-axis direction

  The direction information is defined by a numerical value having a different number of digits for each of the plurality of small beams 8 arranged on each horizontal frame 9A to 9H. In this embodiment, the number of digits of the direction information of each beam 8 is set to be the same as the number of horizontal frame surfaces 9 or frame frames 13 on which the beam 8 can be arranged. Further, each digit of the direction information defines whether the small beam 8 is arranged on the horizontal frame 9 or the small frame 13.

  For example, as shown in FIG. 8, in each horizontal frame surface 9 </ b> A to 9 </ b> H (in this example, the first horizontal frame surface 9 </ b> A is represented), the first beam 8 a is one of the horizontal frame surfaces 9. Can only be placed. Therefore, as shown in FIG. 11 (a), the direction information stored in the direction locus 22a of the first beam 8a consists of a single digit value. Further, the numerical value of one digit of the direction information defines whether or not the first beam 8a is arranged on the horizontal frame 9.

  For example, when the direction information of the first beam 8a is “2”, the first beam 8a is arranged on the horizontal frame 9 with its longitudinal direction parallel to the Y-axis direction. On the other hand, as shown in FIG. 11B, when the direction information of the first beam 8a is “1”, the first beam 8a is parallel to the horizontal frame 9 and the longitudinal direction is parallel to the X-axis direction. Be placed. When the direction information of the first beam 8a is “0”, the first beam 8a is not arranged on the horizontal frame 9. In addition, when the 1st small beam 8 is not arrange | positioned, all the small beams 8 after the 2nd small beam 8 are not arrange | positioned. Therefore, when the direction information of the first beam 8a is “0”, the direction information and position information after the second beam 8b are ignored.

  As shown in FIG. 12A, in each horizontal frame surface 9A to 9H (in this example, the first horizontal frame surface 9A is shown as a representative), the second small beam 8b is defined by the first small beam 8a. It can be arranged on either of the two small frames 13A, 13B that are separated. Therefore, the direction information stored in the direction locus 22a of the second beam 8b consists of a two-digit numerical value. The first digit of the direction information defines whether or not the second small beam 8b is arranged on the first small frame 13A. Further, the numerical value of the second digit of the direction information defines whether or not the second small beam 8b is arranged on the second small frame surface 13B.

  For example, when the direction information is “10”, the numerical value of the first digit is “0 (no beam 8 is arranged)”. Thereby, the 2nd small beam 8b is not arrange | positioned on 13 A of 1st small frame surfaces. On the other hand, the numerical value of the second digit is “1 (parallel to the X-axis direction)”. Thereby, the second small beam 8b is arranged on the second small frame surface 13B with the longitudinal direction parallel to the X-axis direction.

  As shown in FIG. 12B, when the direction information is “02”, the numerical value of the first digit is “2 (Y-axis direction)”. As a result, the second beam 8b is arranged on the first small frame surface 13A with the longitudinal direction parallel to the Y-axis direction. On the other hand, the numerical value of the second digit is “0 (no beam 8 is arranged)”. Thereby, the 2nd small beam 8 is not arrange | positioned on the 2nd small frame surface 13B.

  Further, when the direction information is “00”, both the first digit and the second digit are “0 (no beam 8 is arranged)”. In 13B, the second small beam 8b is not arranged. Further, since only one second beam 8b can be arranged on each horizontal frame 9A to 9H, numerical values other than 0 are set in both the first digit and the second digit of the direction information. There is nothing.

  Similarly, the direction information stored in the direction locus 22a of the third beam 8c is a three-digit numerical value. Further, the numerical value of the first digit defines whether or not the third small beam 8c is arranged on the first small frame surface 13A. Further, the numerical value of the second digit defines whether or not the third small beam 8c is arranged on the second small frame 13B. Also, the numerical value of the third digit defines whether or not the third small beam 8c is arranged on the third frame structure 13C (not shown).

  In this way, each digit of the direction information of the present embodiment defines whether or not the small beam 8 is arranged on the horizontal frame 9 or the small frame 13, so that not only the longitudinal direction of the small beam 8 but also the small number is small. A horizontal frame 9 or a small frame 13 on which the beam 8 is arranged can be defined. Therefore, such direction information is useful for reducing the number of loci 22 of the chromosome information 21.

  Next, the position information defines a fixed position of the reference end portion 18 of the small beam 8 on the horizontal frame 9 or the small frame 13 on which the small beams 8 are arranged. As shown in FIG. 11A and FIG. 12A, the position information of the present embodiment is the length of the side 25 with which the reference end portion 18 abuts on each of the horizontal frame surface 9 or the small frame surface 13. It is defined by the ratio of the distance L1 between the reference points 15a, 15b and the reference end 18 of the beam 8 to L2. Therefore, the numerical value that the position information of the present embodiment can take is 0.0 to 1.0.

  For example, when the direction information of the first beam 8a is “2” and the position information is “0.4”, the fixed position of the reference end 18 of the first beam 8a is determined by the reference end 18 being A distance L1x (X-axis direction) from the reference point 15a of the horizontal frame surface 9 is defined as a node 17 having a length L2x of the side 25 (X-axis direction) that comes into contact. Note that when the numerical value of the position information is “0.5”, for example, the node 17 does not exist on the horizontal frame 9. In such a case, the fixed position of the reference end 18 is defined as the node 17 closest to the numerical value of the position information.

  When the direction information is “0”, the first beam 8 is not arranged on the horizontal frame 9. For this reason, for example, a symbol such as “*” may be input as the position information to distinguish it from other position information.

  As shown in FIG. 12A, when the direction information of the second small beam 8b is “10” and the position information is “0.5”, the reference end portion 18 of the second small beam 8b is fixed. The position is such that the distance L1y (Y-axis direction) from the reference point 15b of the second small frame 13B is 0.5 with respect to the length L2y of the side 25 (Y-axis direction) with which the reference end 18 abuts. It is defined at a certain node 17.

  Thus, the position information of the present embodiment can define the fixed position of the reference end portion 18 of the beam 8. Therefore, the position information is used together with the direction information, so that the arrangement of the small beams 8 can be uniquely defined.

  In step S33 of the present embodiment, the computer 1 converts the direction gene and the position gene into the direction locus 22a and the position locus based on the design constraint condition in each beam 8 of each horizontal frame 9A to 9H. 22b. Thereby, each chromosome information 21 can specify one frame 2 (design sample) in which each of the horizontal beams 9A to 9H is arranged with a number of small beams 8 equal to or less than the upper limit value.

  The direction gene and the position gene are desirably determined at random by the computer 1 according to a random number function, for example. Thereby, since the direction gene and the position gene are set irregularly in the direction locus 22a and the position locus 22b, the chromosome information 21 of various variations can be easily set. Such chromosome information 21 is stored in the computer 1 as numerical data.

  Next, the computer 1 calculates at least one optimum solution of the design factor based on the genetic algorithm (GA) (optimization calculation step S4).

  A genetic algorithm is a learning algorithm that mimics the process by which a living organism adapts to the environment and evolves. In this genetic algorithm, genetic operations such as crossover or mutation are repeated for a plurality of chromosome information expressed by genes. Thereby, in a genetic algorithm, chromosome information can be evolved in time series from a small number of samples, and an optimal solution can be obtained in a short time. FIG. 13 shows a specific processing procedure of the optimization calculation step S4 of the present embodiment.

  In the optimization calculation step S4 of the present embodiment, first, the first target variable and the second target variable of the frame 2 are calculated based on the chromosome information 21 of the group 20 (calculation step S41).

  As shown in FIG. 14, the first target variable of the present embodiment is an approximate cost of the frame 2. This cost is based on, for example, a method of multiplying the weight of the beam set in the chromosome information 21 by a unit price per unit weight of the beam, or a member price stored in a database for each beam. It is calculated by the method of adding up.

  The first target variable is better as the numerical value is smaller. Furthermore, the allowable range of the first target variable is, for example, 1.10 times or less the target cost. The target cost is appropriately set based on, for example, a building budget. Such a first target variable is calculated for each chromosome information 21 and stored in the computer 1.

  The second target variable is the adaptability to the design constraint condition of the frame 2 defined by the direction gene and the position gene. The second target variable and the first target variable are independent variables. This second target variable (fitness to the design constraint conditions) is calculated based on the strength of the frame 2 and the number of direction genes and position genes that violate the design constraint conditions.

  The strength of the frame is obtained, for example, by the retained horizontal strength calculation (so-called route 3 calculation) specified by the Building Standard Law. Further, when there is even one direction gene and position gene that violates the design constraint condition, the second target variable (fitness) is reduced. This 2nd target variable is so good that a numerical value is high, and if it is 1.0 or more, the conditions demanded for frame 2 will be satisfied. Such a second target variable is calculated for each chromosome information 21 (shown in FIG. 12) and stored in the computer 1. When the fitness is less than 1.0, a numerical value may be added as a penalty to the first target variable that is the cost of the frame 2.

  Next, the computer 1 determines whether or not there is at least one chromosome information 21 (hereinafter simply referred to as “optimum solution”) that satisfies both the first target variable and the second target variable (determination). Step S42). In the determination step S42 of the present embodiment, when it is determined that an optimal solution exists among all the chromosome information 21 constituting the group 20, the next manufacturing step S5 is executed.

  On the other hand, if it is determined that there is no optimal solution, the computer 1 performs genetic operations such as crossover and mutation on at least a part of the chromosome information 21 to reconstruct the chromosome information 21 (gene manipulation step). S43). And based on the reconfigure | reconstructed chromosome information 21, calculation process S41 and judgment process S42 are performed again. Thereby, in the optimization calculation process S4, an optimal solution can be obtained reliably.

  In the determination step S42 of the present embodiment, it is determined only whether or not an optimal solution exists, but the present invention is not limited to this. For example, in the determination step S42, in addition to the above conditions, only when the chromosome information 21S with the lowest cost of the frame 2 represented by the first target variable is not updated a plurality of times (for example, 5 to 15 times), The next manufacturing process S5 may be executed.

  In the manufacturing process S5, the frame 2 and the building are manufactured based on the chromosome information 21S having the highest degree of optimization in the group 20 of the last generation. Thereby, in the design method of this embodiment, the frame 2 and the building with the highest strength can be easily and reliably manufactured while keeping the cost of the frame 2 within a predetermined range.

  The chromosome information 21S having the highest degree of optimization is, for example, 1.10 times the first target variable having the lowest first target variable among all the chromosome information 21 having the second target variable of 1 or more. The chromosome information 21 that is the following and has the highest strength of the frame 2 used for the calculation of the second target variable can be determined.

  FIG. 15 shows a specific processing procedure of the gene manipulation step S43 of the present embodiment. In the genetic manipulation step S43 of the present embodiment, first, as shown in FIG. 14, the computer 1 is low in the cost of the frame 2 that represents the plurality of chromosome information 21 belonging to the group 20 by the first target variable. Ranking in order (step S431).

  Next, the computer 1 classifies each chromosome information 21 into an elite group 26 and a non-elite group 27 (step S432). The elite group 26 is composed of chromosome information 21 having a relatively high degree of optimization among all the chromosome information 21 belonging to the group 20. On the other hand, the non-elite group 27 includes chromosome information 21 having a lower degree of optimization than the chromosome information 21 of the elite group 26. The ratio of the elite group 26 can be set as appropriate, but may be determined to be about 5 to 20% of all chromosome information 21 constituting the group 20, for example.

  Next, the computer 1 generates a new population 20 of chromosome information 21 used in the next calculation step S41 (next generation population generation step S433). FIG. 16 shows a specific processing procedure of the next generation group generation step S433 of the present embodiment.

  In the next generation population generation step S433, first, the computer 1 includes the chromosome information 21 of the elite group 26 in the new population 20 without crossover or mutation (step S71). Thereby, in calculation process S41, since the chromosome information 21 of the elite group 26 is included, it can prevent moving away from an optimal solution. The chromosome information 21 of the elite group 26 is stored in the computer 1 as the chromosome information 21 constituting the new group 20.

  Next, the computer 1 performs crossover with respect to a part of chromosome information 21 constituting the group 20 (crossover step S72).

  As shown in FIGS. 17 (a) and 17 (b), in the present embodiment, for example, between each pair of chromosome information 21a and 21b, the gene loci 22a and 22b sandwiched between two crossing points 29 and 29 are provided. The stored direction information or position information is replaced. In such a crossover, direction information or position information defining the same beam 8 can be exchanged between the pair of chromosome information 21a and 21b. Therefore, crossover can be easily reconfigured without setting the direction information stored in the direction locus 22a and the position information stored in the position locus 22b again. The reconstructed chromosome information 21 is stored in the computer 1 as the chromosome information 21 constituting the new group 20. Note that the intersection points 29 and 29 are desirably set at random by the computer 1.

  In the present embodiment, the case where the crossing is a two-point crossing in which the direction information and the position information sandwiched between the two crossing points 29 and 29 are exchanged is exemplified, but the present invention is not limited to this. As the crossover, for example, a single point crossover, a multipoint crossover, or a uniform crossover may be used, or a combination of these may be implemented.

  Next, the computer 1 performs a mutation on a part of chromosome information 21 constituting the group 20 (mutation step S73). In the mutation step S73 of the present embodiment, mutation is performed on the chromosome information 21 that is not subject to crossover.

  As shown in FIG. 18, in the present embodiment, first, in each chromosome information 21, the direction locus 22 a or the position locus 22 b is randomly selected. Next, the direction information and the position information stored in the selected direction locus 22a or the position locus 22b are all that the direction information corresponding to the direction locus 22a or the position locus 22b or the position information can take. Of the numerical values, it is replaced with a randomly selected numerical value.

  Unlike crossover, such mutation is not limited to the direction information or position information defined in each chromosome information 21 constituting the group 20, but is reconstructed with new direction information or position information. can do. Therefore, the mutation can prevent falling into a local optimal solution. The reconstructed chromosome information 21 is stored in the computer 1 as the chromosome information 21 constituting the new group 20.

  Thus, in the optimization calculation step S4 of the present embodiment, the remaining chromosome information 21 is reconstructed while leaving the chromosome information 21 of the elite group 26 having a high degree of optimization of the first target variable and the second target variable. You can try new evolution. Thereby, in the design method of this invention, a design factor can be evolved with few samples. Therefore, the design method of the present invention can find the optimum solution in a short time.

  The proportion of chromosome information 21 to be mutated can be set as appropriate, but may be determined to be, for example, about 10 to 40% of all chromosome information 21 constituting the group 20. If the ratio of chromosome information 21 to be mutated is less than 10%, there is a possibility that gene information cannot be sufficiently evolved. On the other hand, if the ratio of chromosome information 21 to be mutated exceeds 40%, gene information that violates the design constraint condition is greatly increased, and an optimal solution may not be obtained in a short time.

  In addition, after the chromosome information 21 that satisfies the cost of the frame 2 represented by the first target variable and falls within the allowable range and satisfies the fitness represented by the second target variable appears, the vicinity of the optimal solution is generally determined. It can be determined that the search has been completed. For this reason, in the next generation population generation step S433, it is desirable to prioritize the evolution by crossing the favorable chromosome information 21. Thereby, in optimization calculation process S4, an optimal solution can be calculated | required in a short time. In this case, the mutation rate may be set to about 5 to 10% of all chromosome information 21 constituting the group 20.

  The number of chromosome information 21 belonging to the group 20 is preferably 15 to 50. If the number of chromosome information 21 is less than 15, there is a possibility that gene information cannot be sufficiently evolved. Conversely, if the number of chromosome information 21 exceeds 50, a lot of calculation time may be required.

Furthermore, it is desirable that the number of chromosome information 21 belonging to the group 20 is determined by, for example, the number of floors and the total floor area of a building. For example, in the case of a two-story building such as this embodiment, it is desirable that the chromosome information 21 is set according to the number set for each total floor area described below. In the case of a three-story building, it is desirable to set the chromosome information 21 for the number obtained by adding 5 to the following number. Furthermore, in the case of a one-story building, it is desirable to set chromosome information 21 for the number obtained by subtracting 5 from the following number.
100m less than ^2: 20 100~120m ^2: 25 pieces 120~140m ^2: 30 pieces 140~160m ^2: 35 pieces 160~180m ^2: 40 pieces 180~200m ^2: 45 pieces 200 meters ² or more: 50

  In the design method of the present embodiment, as shown in FIG. 9, the gene locus 22 defining the arrangement of each beam 8 is exemplified by only the directional gene locus 22a and the position gene locus 22b. However, the present invention is not limited to this. For example, as shown in FIG. 19, the gene locus 22 may further include a frame surface gene locus 22c in addition to the direction gene locus 22a and the position gene locus 22b.

  The frame plane gene locus 22c is defined in the same number as the upper limit value of the number of the small beams 8 in each horizontal frame plane 9A to 9H, similarly to the direction gene locus 22a and the position gene locus 22b. Therefore, each chromosome information 21 is assigned a frame plane locus 22c for each beam 8 in each of the horizontal frame planes 9A to 9H.

Further, frame surface information is stored in the frame surface locus 22c. The frame surface information specifies the horizontal frame surface 9 or the small frame surface 13 on which the small beams 8 are arranged. This frame surface information is defined by a numerical value of at least one digit, in this embodiment, one digit. The frame surface information of this embodiment is as follows, for example.
0: No arrangement of small beams 1: Horizontal frame or first frame surface 2: Second frame surface 3: Third frame surface

  For example, as shown in FIG. 20, when the frame information of the first beam 8 a is “1”, the first beam 8 a is arranged on the horizontal frame 9. On the other hand, when the frame surface information of the first beam 8 a is “0”, the first beam 8 a is not arranged on the horizontal frame 9.

  As shown in FIG. 21A, when the frame surface information of the second beam 8b is “1”, the second beam 8b is arranged on the first beam surface 13A. In addition, as shown in FIG. 21B, when the frame surface information of the second beam 8b is “2”, the second beam 8b is arranged on the second beam surface 13B. Furthermore, when the frame surface information of the second beam 8b is “0”, the second beam 8b is not disposed on the first beam frame 13A and the second beam frame 13B.

  Thus, since the frame surface information can specify the horizontal frame surface 9 or the frame frame 13 on which each beam 8 is arranged, the number of digits of the direction information is set to the beam 8 as in the previous embodiment. It is not necessary to define the presence or absence of the small beams 8 arranged on the horizontal frame 9 or the small frame 13 by setting the same as the number of the horizontal frame 9 or the small frame 13 that can be arranged. Therefore, since the chromosome information 21 can define the direction information by a single digit value that defines the longitudinal direction of the beam 8, the data structure and the processing procedure can be simplified.

  As shown in FIG. 22A, the locus 22 may further include a number locus 22d. This number locus 22d is defined for each horizontal frame 9A to 9H in each chromosome information 21.

  Each number locus 22d stores number information consisting of a numerical value of at least one digit. This number information defines the number of small beams 8 to be calculated and crossed and mutated in each horizontal frame plane 9A to 9H for the first target variable and the second target variable.

  For example, when the number information of the first horizontal frame surface 9A is the upper limit value (for example, 3) of the number of small beams 8, the first target is set for all the small beams 8 on the first horizontal frame surface 9A. Variables and second target variables are calculated and crossed and mutated.

  On the other hand, as shown in FIG. 22B, when the number information of the first horizontal frame 9A is “2”, the first beam 8a and the second beam 8b on the first horizontal frame 9A. Only the third beam 8c is untargeted. Accordingly, the number information can dynamically change the calculation of the first target variable and the second target variable, and the cross beam and the mutating beam 8 in each horizontal frame 9A to 9H during the calculation. .

  Thereby, in this embodiment, for example, in the middle of obtaining the optimum solution, the number of the small beams 8 can be narrowed down based on the results of the first target variable and the second target variable. Can be sought.

2 Frame 3 Column 4 Beam

Claims (14)

A building frame including a column and a beam, wherein the beam includes a plurality of large beams that horizontally connect between the columns, and a small beam that further partitions a horizontal frame surrounded by the large beams. A method for designing using
The frame design method comprising: an optimization calculation step in which the computer calculates at least one optimum solution of design factors for at least one of the small beams based on a genetic algorithm.   The frame design method according to claim 1, wherein the design factor includes an arrangement or a cross-sectional shape of the beam.   3. The frame design method according to claim 1, wherein the optimal solution is the design factor that satisfies both the first target variable and the second target variable independent of each other that can be calculated from the frame. 4. Inputting basic building information including the shape of the building and load conditions to the computer; and
Inputting into the computer design constraints for arranging the beam on the frame;
The computer includes a group generation step of generating a group consisting of a plurality of types of chromosome information with different design factors based on the design constraint condition,
The frame design method according to claim 1, wherein the optimization calculation step calculates the optimal solution using the group. The optimization calculation process includes:
A calculation step of calculating a first target variable and a second target variable of each frame based on each chromosome information;
A gene manipulation step of crossing and mutating at least a part of the chromosome information, and determining whether or not there is at least one chromosome information satisfying both the first target variable and the second target variable The frame structure designing method according to claim 4, further comprising a step. The group includes an elite group consisting of chromosome information with a high degree of optimization of the first target variable and the second target variable,
6. The frame design method according to claim 5, wherein, in the gene manipulation step, the chromosome information of the elite group is included in the new group without being crossed and mutated. The chromosome information includes at least one genetic locus capable of storing genetic information defining the arrangement of one trabecule for each horizontal frame.
The genetic information includes direction information that defines a longitudinal direction of the beam, and position information that defines a fixed position of the beam.
The locus includes a direction locus capable of storing the direction information, and a position locus capable of storing the position information,
The frame structure design method according to any one of claims 4 to 6, wherein the chromosome information includes the directional locus and the position locus for each beam on each horizontal frame surface. In the design constraint condition, the direction and position where the beam can be fixed are set,
8. The frame according to claim 7, wherein the group generation step includes a step of storing the direction information and the position information determined by the computer based on the design constraint condition in the direction locus and the position locus. Body design method. As the design constraint condition, an upper limit value of the number of the small beams arranged on each horizontal frame is set,
The upper limit is smaller than the maximum number of the small beams that can be arranged on each horizontal frame surface,
9. The frame structure design method according to claim 7 or 8, wherein the chromosome information is defined such that the number of the directional loci and the position loci is the same as the upper limit value in each horizontal frame plane. Each horizontal frame is divided into a plurality of small frames by the beam.
The beam is disposed on the horizontal frame or the small frame,
The direction information of each beam is composed of letters or numerical values having the same number of digits as the number of horizontal frame surfaces or small frame surfaces on which the beam can be arranged,
10. Each of the digits of the direction information is defined by the character or the numerical value stored in each digit as to whether or not the small beam is arranged on the horizontal frame or the small frame. A frame design method according to any one of the above. Each horizontal frame is divided into a plurality of small frames by the beam.
The beam is arranged on the horizontal frame surface or the small frame surface,
The gene information further includes frame surface information for specifying the horizontal frame surface or the frame surface on which the beam is arranged,
The frame design method according to any one of claims 7 to 9, wherein the gene locus further includes a frame surface gene locus capable of storing the frame surface information. The optimization calculation step includes a calculation step of calculating a first target variable and a second target variable of each frame based on each chromosome information, and crossover and mutation with respect to at least a part of the chromosome information. Including a genetic manipulation step
The genetic information includes, in each horizontal frame plane, calculation of the first target variable and the second target variable, and number information that defines the number of the cross beams to be crossed and mutated,
The locus further includes a number locus capable of storing the number information,
Each chromosome information is defined by one number locus,
In the calculation step and the gene manipulation step, the first target variable and the second target variable are set only for the gene loci of the beam corresponding to the number set in the number information in each horizontal frame. The method of designing a frame structure according to any one of claims 7 to 11, wherein calculation is performed and crossover and mutation are performed.   A building manufacturing method for manufacturing the frame and the building based on at least one of the optimum solutions calculated by the frame designing method according to any one of claims 1 to 12.   A design apparatus including an arithmetic processing unit for executing the frame design method according to claim 1.

Images