JP7649657B2 - Building frame estimation device and building frame estimation method - Google Patents

Description

The present invention relates to a building frame estimation device and a building frame estimation method.

Conventionally, in the early stages of structural design of a building, when determining the structural frame from the building volume, which is the volume of the building's frame, the designer makes the decision individually, incorporating engineering judgment based on past experience and knowledge, past similar cases, etc. Note that the technology described in Patent Document 1 is known as a technology for designing building frames.

JP 2016-110298 A

However, for example, if the designer is inexperienced, they may lack past experience and knowledge, which can lead to rework in the next design process, resulting in inconsistent design accuracy for the structural frame, and making it difficult to determine an appropriate structural frame based on the building's structural volume.

The present invention has been made to solve the above problems, and its purpose is to provide a building frame estimation device and a building frame estimation method that can determine an appropriate structural frame from the frame volume of a building.

In order to solve the above problem, one aspect of the present invention provides a method for estimating the building frame of a building from a joint number vector inner product value, which is the inner product value of a first joint number vector, which is a vector whose elements are the numbers of each type of joint connecting columns and beams of the building in a first direction parallel to a horizontal plane of the building, and a second joint number vector, which is a vector whose elements are the numbers of each type of joint in a second direction parallel to the horizontal plane and different from the first direction, the frame volume of the building , and basic design quantities taken into consideration in structural design. and a building frame estimation unit that estimates a building frame of the building from the building frame estimation model stored in the model storage unit, the joint number vector inner product value, the frame volume of the building , and the basic design quantities, wherein the building frame estimation model is constructed based on learning data having a plurality of sets of data including at least the joint number vector inner product value, the frame volume, the basic design quantities, and the building frame of the building .

According to another aspect of the present invention, there is provided a method for estimating a building frame of the building from a first joint number vector, the first joint number vector being a vector having elements each representing the number of each type of joint connecting a column and a beam of the building in a first direction parallel to a horizontal plane of the building, and the first joint number vector being a vector having elements each representing the number of each type of joint in a second direction parallel to the horizontal plane and different from the first direction, a frame volume of the building, and a model storage unit that stores a building frame estimation model for estimating a building frame of the building from the building frame volume and basic design quantities taken into consideration in structural design ; and a building frame estimation unit , wherein the building frame estimation unit includes a building frame estimation step of estimating a building frame of the building from the building frame estimation model stored in the model storage unit, the joint number vector dot product value, the frame volume of the building, and the basic design quantities, and the building frame estimation model is constructed based on learning data having a plurality of sets of data including at least the joint number vector dot product value, the frame volume, the basic design quantities, and the building frame of the building .

According to the present invention, it is possible to determine an appropriate structural frame based on the building's frame volume.

FIG. 1 is a block diagram showing an example of a building frame estimation device according to an embodiment of the present invention. 1 is a diagram illustrating an example of a building frame and a joint in this embodiment. FIG. 10A to 10C are diagrams illustrating an example of a process for constructing a building frame estimation model in this embodiment. 10 is a flowchart showing an example of a building frame estimation model construction process of the building frame estimation device according to the present embodiment. 11 is a flowchart showing an example of a building frame estimation process of the building frame estimation device according to the present embodiment.

The building frame estimation device and building frame estimation method according to one embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an example of a building frame estimation device 1 according to this embodiment.
1, the building frame estimation device 1 is, for example, a computer device such as a personal computer or a server device. The building frame estimation device 1 acquires, for example, a frame volume of a building and basic design quantities considered in structural design (for example, a reference floor area, an XY span length, and a floor height) as input data, estimates a building frame of a building based on the input data, and outputs the building frame.
The building frame estimation device 1 includes an input unit 11, a display unit 12, a memory unit 13, and a control unit 14.

The input unit 11 is an input device such as a keyboard or a mouse, and receives various information such as input data, and outputs it to the control unit 14 .
The display unit 12 is, for example, a display device such as a liquid crystal display, and displays various information such as the building frame of the building, which is output data (output information). The display unit 12 is an example of an output unit that outputs output information. When using software, the building frame estimation device 1 displays the building frame of the building on the display unit 12 using a three-dimensional modeling tool or the like.

The memory unit 13 stores various information used by the building frame estimation device 1. The memory unit 13 includes a learning data memory unit 131, a model memory unit 132, an input information memory unit 133, an inner product information memory unit 134, and a building frame memory unit 135.

The learning data storage unit 131 stores learning data for constructing a building frame estimation model, which will be described later. The learning data storage unit 131 stores learning data having multiple sets of data that associate, for example, the building frame volume, the area of the building's reference floor (reference floor area), the span length in the X direction, the span length in the Y direction, the floor height, the joint number vector inner product value, which will be described later, and the building frame. Here, the building frame and the joint number vector inner product value will be described.

FIG. 2 is a diagram illustrating an example of a building frame and a joint in this embodiment.
Fig. 2(a) shows an example of a building frame. In Fig. 2(a), directions parallel to the bottom surface (horizontal surface) of a reference floor of the building are the X direction and the Y direction, and the X direction and the Y direction are orthogonal or nearly orthogonal to each other.

As shown in FIG. 2(a), the building frame has a structure in which columns and beams are joined together, and is divided into three sections: the top floor, the middle floor, and the bottom floor.
FIG. 2(b) shows a cross section of another example of a building frame.
As shown in Fig. 2(b), in the building frame, columns and beams are joined by four types of joints, including an L-shaped joint K1 (first joint), a T-shaped joint K2 (second joint), a T-shaped joint K3 (third joint), and a cross-shaped joint K4 (fourth joint).

The L-shaped joint K1 is an L-shaped joint that connects a column and a beam at the corner of a building. The T-shaped joint K2 is a joint that connects a column and a beam on the periphery of a building. The T-shaped joint K3 is a T-shaped joint that connects a beam of the building to a column on either the top or bottom of the beam. The cross-shaped joint K4 is a cross-shaped joint that connects a column and a beam of the building in a cross shape.

In this embodiment, a vector whose elements are the number of L-shaped joints K1, T-shaped joints K2, T-shaped joints K3, and cross-shaped joints K4 in the X direction is defined as the joint number vector in the X direction (first joint number vector). Also, a vector whose elements are the number of L-shaped joints K1, T-shaped joints K2, T-shaped joints K3, and cross-shaped joints K4 in the Y direction is defined as the joint number vector in the Y direction (second joint number vector).

Furthermore, a vector (junction number vector) having the numbers of the above-mentioned four types of junctions as elements is generated for each of the top floor, middle floor, and bottom floor.
In this embodiment, the inner product value of the junction number vector in the X direction and the junction number vector in the Y direction is set as the junction number vector inner product value.

Here, the joint number vector UVx in the X direction, the joint number vector UVy in the Y direction, and the joint number vector inner product value UIP on the top floor are expressed by the following formula (1).

Here, ULx, UTx, UWx, and UPx indicate the number of L-shaped joints K1, T-shaped joints K3, T-shaped joints K2, and cross-shaped joints K4 in the X direction of the top floor, while ULy, UTy, UWy, and UPy indicate the number of L-shaped joints K1, T-shaped joints K3, T-shaped joints K2, and cross-shaped joints K4 in the Y direction of the top floor.
In addition, there are no T-shaped joints K2 and cross-shaped joints K4 on the top floor of the building, and the number thereof is "0".

Furthermore, the joint number vector MVx in the X direction, the joint number vector MVy in the Y direction, and the joint number vector inner product value MIP on the intermediate floor are expressed by the following formula (2).

Here, MLx, MTx, MWx, and MPx indicate the number of L-shaped joints K1, the number of T-shaped joints K3, the number of T-shaped joints K2, and the number of cross-shaped joints K4 in the X direction of the intermediate floor. Also, MLy, MTy, MWy, and MPy indicate the number of L-shaped joints K1, the number of T-shaped joints K3, the number of T-shaped joints K2, and the number of cross-shaped joints K4 in the Y direction of the intermediate floor.

Furthermore, the joint number vector BVx in the X direction, the joint number vector BVy in the Y direction, and the joint number vector inner product value BIP on the bottom floor are expressed by the following formula (3).

Here, BLx, BTx, BWx, and BPx indicate the number of L-shaped joints K1, T-shaped joints K3, T-shaped joints K2, and cross-shaped joints K4 in the X direction of the lowest floor, while BLy, BTy, BWy, and BPy indicate the number of L-shaped joints K1, T-shaped joints K3, T-shaped joints K2, and cross-shaped joints K4 in the Y direction of the lowest floor.
In addition, there are no T-shaped joints K2 and cross-shaped joints K4 on the bottom floor of the building, and the number thereof is "0".

Returning to the explanation of FIG. 1, the learning data includes the above-mentioned joint number vector inner product values (the inner product value UIP of the top floor, the inner product value MIP of the middle floor, and the inner product value BIP of the bottom floor). In addition, the building frame included in the learning data may be generated, for example, using an existing 3D modeling tool or the like based on the area of the building's reference floor (reference floor area), the span length in the X direction, the span length in the Y direction, and the floor height, or may be previously designed. In other words, each set of data in the learning data is data generated using an existing 3D modeling tool or the like according to various shapes, frame volumes, and the like, or data previously designed.

The model storage unit 132 stores a building frame estimation model that estimates the building frame of a building from the joint number vector inner product value, the building frame volume, and basic design quantities considered in structural design. The building frame estimation model is a model constructed based on the learning data stored in the learning data storage unit 131 described above. The building frame estimation model is a model that classifies the building frame into multiple groups by machine learning such as clustering using the frame volume and the joint number vector inner product value as features, and outputs labels (group identification information) indicating the classified groups from the frame volume and the joint number vector inner product value, and details of this will be described later.

The input information storage unit 133 stores various information (e.g., standard floor area, span length in the X direction, span length in the Y direction, and floor height) input from outside when estimating the building structure.

The inner product information storage unit 134 stores the inner product values of the joint number vector (the inner product value UIP of the top floor, the inner product value MIP of the middle floor, and the inner product value BIP of the bottom floor) used as input data when estimating the building frame. The inner product information storage unit 134 may also store the joint number vector in the X direction and the joint number vector in the Y direction.

The building structure storage unit 135 is, for example, a database of building structures, and stores the above-mentioned labels (group identification information) in association with building structures. The building structures stored in the building structure storage unit 135 are, for example, building structures used to construct a building structure estimation model, building structures of buildings designed in the past, etc. The building structure storage unit 135 may be configured to add building structures of buildings designed in the past as know-how from time to time.

The control unit 14 is a processor including, for example, a CPU (Central Processing Unit) and generally controls the building frame estimation device 1. The control unit 14 mainly executes the construction process of a building frame estimation model and the estimation process of the building frame. The control unit 14 includes a model construction unit 141, a data acquisition unit 142, an inner product generation unit 143, a building frame estimation unit 144, and a display control unit 145.

The model construction unit 141 performs machine learning based on the learning data stored in the learning data storage unit 131 to construct a building frame estimation model. The model construction unit 141 stores the constructed building frame estimation model in the model storage unit 132. Here, the construction process of the building frame estimation model by the model construction unit 141 will be described with reference to FIG. 3.

FIG. 3 is a diagram illustrating an example of a process for constructing a building frame estimation model in this embodiment.
3A shows a multidimensional chart in which each group of learning data stored in the learning data storage unit 131 is classified by machine learning such as clustering using the frame volume and the inner product value of the joint number vector as feature values. In FIG. 3A, the dashed line portion F1 corresponds to the frame volume and the inner product value of the joint number vector of the building.

The model construction unit 141 executes a clustering process as shown in FIG. 3(a) to classify the building into groups (G1, G2, G3, ...), and assigns a label to each group as group identification information. The model construction unit 141 classifies the building frame of a building into multiple groups by a clustering process using the frame volume and the joint number vector inner product value as features.

Next, the model construction unit 141 constructs an estimation model that estimates each group (group identification information) classified by clustering processing from the frame volume and joint number vector inner product value of the learning data, for example, by deep learning such as a neural network. The model construction unit 141 constructs, for example, a neural network NN1 as shown in Fig. 3(b) as a building frame estimation model by machine learning. The neural network NN1 estimates a group of building frames (a label assigned to the group) from the frame volume, basic design quantities considered in structural design, and the joint number vector inner product value.
The model construction unit 141 stores the constructed building frame estimation model in the model storage unit 132.

Returning to the explanation of FIG. 1, the data acquisition unit 142 acquires input data (input information) when estimating the building frame. For example, the data acquisition unit 142 acquires, via the input unit 11, as input data, the frame volume, the area of the reference floor of the building (reference floor area), the span length in the X direction, the span length in the Y direction, and the floor height indicating the height of the floor of the building. The data acquisition unit 142 stores the acquired input data in the input information storage unit 133.

The inner product generation unit 143 generates a junction number vector inner product value based on the area of the reference floor, the span length in the X direction, the span length in the Y direction, and the floor height acquired by the data acquisition unit. The inner product generation unit 143 calculates the number of each type of junction based on the input data stored in the input information storage unit 133, and generates the junction number vector inner product values (the inner product value UIP of the top floor, the inner product value MIP of the intermediate floors, and the inner product value BIP of the bottom floor) using the above-mentioned formulas (1) to (3). The inner product generation unit 143 stores the generated junction number vector inner product values (the inner product value UIP of the top floor, the inner product value MIP of the intermediate floors, and the inner product value BIP of the bottom floor) in the inner product information storage unit 134.

The building frame estimation unit 144 estimates the building frame of the building from the building frame estimation model stored in the model storage unit 132, the joint number vector dot product value, the building frame volume, and basic design quantities considered in structural design. The building frame estimation unit 144 estimates the building frame of the building based on, for example, the dot product value of the top floor, the dot product value of the middle floor, and the dot product value of the bottom floor stored in the dot product information storage unit 134, and the building frame volume and basic design quantities considered in structural design stored in the input information storage unit 133.

Specifically, the building structure estimation unit 144 acquires group identification information based on the building structure estimation model, and estimates the building structure belonging to the group corresponding to the acquired label (group identification information) as the building structure. For example, the building structure estimation unit 144 acquires the building structure corresponding to the acquired label from the building structure storage unit 135, and outputs the acquired building structure as the building structure of the estimation result.

The display control unit 145 (an example of an output control unit) outputs output information indicating the building structure of the building estimated by the building structure estimation unit 144 to the display unit 12 (output unit). That is, the display control unit 145 causes, for example, a three-dimensional model or image of the building structure to be displayed on the display unit 12. The display control unit 145 causes the display unit 12 to display multiple building structures belonging to a group corresponding to the label (group identification information) estimated by the building structure estimation unit 144. In this case, the display control unit 145 may, for example, preferentially display on the display unit 12 a building structure that is close to the reference building structure (average building structure) of the estimated group.

Next, the operation of the building frame estimation device 1 according to this embodiment will be described with reference to the drawings.
FIG. 4 is a flowchart showing an example of a building structure estimation model construction process of the building structure estimation device 1 according to this embodiment.

As shown in FIG. 4, in the process of constructing a building frame estimation model, the building frame estimation device 1 first acquires learning data (step S101). The model construction unit 141 of the building frame estimation device 1 acquires the learning data stored in the learning data storage unit 131.

Next, the model construction unit 141 classifies the building structures by clustering processing or the like using the structure volume and the inner product of the joint number vector as features, and assigns labels (step S102). The model construction unit 141 executes a clustering process as shown in FIG. 3(a) to classify the structures into groups (G1, G2, G3, ...), and assigns a label to each classified group as group identification information.

Next, the model construction unit 141 constructs an estimation model that estimates a label from the frame volume, basic design quantities, and the dot product value of the joint number vector (step S103). The model construction unit 141 uses machine learning to construct, for example, a neural network NN1 as shown in FIG. 3(b) as a building frame estimation model. The neural network NN1 takes the basic design quantities considered in structural design as input values, and estimates the group of building frames (the label to be assigned to the group) using the frame volume and the dot product value of the joint number vector as judgment criteria.

Next, the model construction unit 141 stores the constructed estimation model in the model storage unit 132 as a building frame estimation model (step S104). After the processing of step S104, the model construction unit 141 ends the construction process of the building frame estimation model.

Next, the building frame estimation process of the building frame estimation device 1 according to this embodiment will be described with reference to FIG. 5 .
FIG. 5 is a flowchart showing an example of a building frame estimation process of the building frame estimation device 1 according to this embodiment.

As shown in FIG. 5, in the building frame estimation process, the building frame estimation device 1 first acquires input data (step S201). The data acquisition unit 142 of the building frame estimation device 1 acquires, for example, via the input unit 11, the frame volume, the area of the reference floor of the building, the span length in the X direction, the span length in the Y direction, and the floor height indicating the height of the floor of the building as input data. The data acquisition unit 142 stores the acquired input data in the input information storage unit 133.

Next, the inner product generation unit 143 of the building frame estimation device 1 generates a joint number vector inner product value from the input data (step S202). The inner product generation unit 143 calculates the number of each type of joint based on the input data stored in the input information storage unit 133, and generates the joint number vector inner product values (the inner product value UIP of the top floor, the inner product value MIP of the middle floor, and the inner product value BIP of the bottom floor) using the above-mentioned formulas (1) to (3). The inner product generation unit 143 stores the generated joint number vector inner product values in the inner product information storage unit 134.

Next, the building frame estimation unit 144 of the building frame estimation device 1 estimates the label of the group from the building frame estimation model and the frame volume of the building and the joint number vector inner product value, etc. (step S203). The joint number vector inner product value, etc. includes, for example, basic design quantities to be considered in structural design. The building frame estimation unit 144 estimates the label of the group using the building frame estimation model (neural network NN1) stored in the model storage unit 132 from the inner product value of the top floor, the inner product value of the middle floor, and the inner product value of the bottom floor stored in the inner product information storage unit 134, and the frame volume of the building and basic design quantities to be considered in structural design stored in the input information storage unit 133.

Next, the building frame estimation device 1 outputs the building frame of the group corresponding to the label (step S204). The building frame estimation unit 144 acquires the building frame corresponding to the estimated label from the building frame storage unit 135. The display control unit 145 of the building frame estimation device 1 causes, for example, a three-dimensional model or image of the acquired building frame to be displayed on the display unit 12. After processing of step S204, the building frame estimation device 1 ends the building frame estimation process.

As described above, the building frame estimation device 1 according to the present embodiment includes a model storage unit 132 and a building frame estimation unit 144. The model storage unit 132 is a building frame estimation model that estimates the building frame of a building from the joint number vector inner product value and the frame volume of the building, and stores the building frame estimation model constructed based on learning data having a plurality of set data including at least the joint number vector inner product value, the frame volume, and the building frame of the building. Here, the joint number vector inner product value is the inner product value of the X-direction joint number vector (first joint number vector), which is a vector whose elements are the number of each type of joint that connects the columns and beams of the building in the X-direction (first direction) parallel to the horizontal plane of the object, and the Y-direction joint number vector (second joint number vector), which is a vector whose elements are the number of each type of joint in the Y-direction (second direction different from the first direction) that is parallel to the horizontal plane and perpendicular or nearly perpendicular to the X-direction. The building frame estimation unit 144 estimates the building frame of the building from the building frame estimation model stored in the model storage unit 132, the joint number vector inner product value, and the frame volume of the building.

As a result, the building frame estimation device 1 according to this embodiment estimates suitable building frame candidates from the building frame volume and the joint number vector inner product value, so that even if the designer has little experience, for example, it is possible to determine an appropriate structural frame. In addition, the building frame estimation device 1 according to this embodiment uses the inner product value of the joint number vector in the X direction and the joint number vector in the Y direction as a feature value, so that the feature value can be reduced and processing can be simplified compared to when the joint number vector is used as it is as a feature value. For example, the joint number vector inner product value UIP in the above-mentioned formula (1), the joint number vector inner product value MIP in the formula (2), and the joint number vector inner product value BIP in the formula (3) each reduce a total of eight feature values of the XY directions and the four patterns of joints to a one-dimensional numerical value.

In addition, in this embodiment, the joint number vector inner product value includes the inner product value UIP of the top floor of the building, the inner product value MIP of the middle floor of the building, and the inner product value BIP of the bottom floor of the building. The building frame estimation unit 144 estimates the building frame of the building based on the inner product value UIP of the top floor, the inner product value MIP of the middle floor, and the inner product value BIP of the bottom floor, and the frame volume.

As a result, the building frame estimation device 1 according to this embodiment can estimate a more appropriate building frame according to the shape by dividing the dot product value UIP of the top floor, the dot product value MIP of the middle floor, and the dot product value BIP of the bottom floor, since the types and numbers of joints tend to differ on the top floor, middle floors, and bottom floor depending on the shape of the building.

In addition, in this embodiment, the building frame estimation model classifies the building frames into multiple groups by clustering based on the frame volume and the joint number vector inner product value, and outputs a label (group identification information) indicating the classified group from the frame volume and the joint number vector inner product value. The building frame estimation unit 144 acquires the group identification information based on the building frame estimation model, and estimates the building frames belonging to the group corresponding to the acquired label (group identification information) as the building frames of the building.

As a result, the building structure estimation device 1 according to this embodiment estimates groups classified by clustering, allowing the user to select an appropriate building structure from among multiple candidates belonging to a group.

The building frame estimation device 1 according to this embodiment also includes a data acquisition unit 142, an inner product generation unit 143, and a display control unit 145 (output control unit). The data acquisition unit 142 acquires the frame volume, the area of the reference floor of the building, the span length in the X direction, the span length in the Y direction, and the floor height indicating the height of the floor of the building. The inner product generation unit 143 generates an inner product value of the joint number vector based on the area of the reference floor, the span length in the X direction, the span length in the Y direction, and the floor height acquired by the data acquisition unit 142. The display control unit 145 outputs output information indicating the building frame of the building estimated by the building frame estimation unit 144 to the display unit 12 (output unit).

As a result, the building frame estimation device 1 according to this embodiment calculates the joint number vector inner product value from the input data, which are the area of the reference floor of the building, the span length in the X direction, the span length in the Y direction, and the floor height of the building, so that the user does not need to generate the joint number vector inner product value, thereby improving convenience. Also, in the building frame estimation device 1 according to this embodiment, the display control unit 145 outputs output information indicating the building frame of the estimated building to the display unit 12 (output unit), so that the user can, for example, visually confirm the building frame, making it easier for the user to select a more appropriate building frame.

In this embodiment, the types of joints include an L-shaped joint (first joint) that connects a column and a beam at the corner of a building, a T-shaped joint (second joint) that connects a column and a beam on the periphery of the building, a T-shaped joint (third joint) that connects a beam of the building to a column on either the top or bottom of the beam, and a cross-shaped joint (fourth joint) that connects a column and a beam of the building in a cross shape. The joint count vector in the X direction and the joint count vector in the Y direction include the number of L-shaped joints K1, T-shaped joints K2, T-shaped joints K3, and cross-shaped joints K4 as vector elements.

As a result, the building frame estimation device 1 according to this embodiment is able to appropriately capture the characteristics of a building by using a joint number vector, since the type and number of joints tend to vary depending on the shape of the building.

Moreover, the building frame estimation device 1 according to this embodiment includes a model construction unit 141 that constructs a building frame estimation model based on the learning data.
As a result, the building frame estimation device 1 according to this embodiment can reconstruct a building frame estimation model with updated learning data, for example, periodically.

The building frame estimation method according to the present embodiment is a building frame estimation method of the building frame estimation device 1 having the model storage unit 132, and includes a building frame estimation step. The model storage unit 132 stores a building frame estimation model that estimates the building frame of the building from a joint number vector inner product value and a frame volume of the building, the building frame estimation model being constructed based on learning data having a plurality of sets of data including at least the joint number vector inner product value, the frame volume, and the building frame of the building. In the building frame estimation step, the building frame estimation device 1 estimates the building frame of the building from the building frame estimation model stored in the model storage unit 132, the joint number vector inner product value, and the frame volume of the building.
As a result, the building frame estimation method according to this embodiment has the same effects as the building frame estimation device 1 described above, and can determine an appropriate structural frame even if the designer has little experience, for example.

The present invention is not limited to the above-described embodiment, and can be modified without departing from the spirit of the present invention.
For example, in the above embodiment, a configuration example in which the building frame estimation device 1 includes the model construction unit 141 has been described, but the present invention is not limited to this, and the building frame estimation model may be constructed outside the building frame estimation device 1. In this case, for example, the model construction device may include the model construction unit 141 and construct the building frame estimation model.

In the above embodiment, the model construction unit 141 uses clustering to construct a model that estimates the labels (group identification information) of classified groups, but the present invention is not limited to this example. The model construction unit 141 may construct a model that directly estimates the building structure instead of the group labels (group identification information).

In the above embodiment, the building frame estimation device 1 is configured as a single device, but the present invention is not limited to this and may be configured as a plurality of devices (e.g., a plurality of server devices). In addition, the building frame estimation device 1 may have a part or all of the memory unit 13 provided externally.

In addition, in the above embodiment, an example was described in which the output unit was the display unit 12, but this is not limited to this and the output unit may be another output unit, for example, a printer or a storage device such as a file server.
In the above embodiment, an example of the second direction different from the first direction is described in which two directions (X direction and Y direction) that are perpendicular or nearly perpendicular to each other are used; however, the present invention is not limited to this example, and different directions having another relationship may also be used.

In addition, in the above embodiment, an example of using clustering has been described as an example of a machine learning method for classifying building structures into multiple groups, but this is not limited to this, and other methods such as dimensionality reduction and principal component analysis may also be used.

Each component of the building frame estimation device 1 described above has an internal computer system. A program for realizing the functions of each component of the building frame estimation device 1 described above may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed to perform processing in each component of the building frame estimation device 1 described above. Here, "reading a program recorded on a recording medium into a computer system and executing it" includes installing the program into a computer system. Here, "computer system" includes hardware such as an OS and peripheral devices.

A "computer system" may also include multiple computer devices connected via a network, including the Internet, a WAN, a LAN, a dedicated line, or a virtual desktop communication line. A "computer-readable recording medium" refers to portable media such as a flexible disk, an optical magnetic disk, a ROM, a CD-ROM, or a storage device such as a hard disk built into a computer system. In this way, the recording medium storing the program may be a non-transitory recording medium such as a CD-ROM.

The recording medium also includes an internal or external recording medium accessible from a distribution server to distribute the program. The program may be divided into multiple parts, downloaded at different times, and then combined by each component of the building frame estimation device 1, or each divided program may be distributed by a different distribution server. Furthermore, the term "computer-readable recording medium" includes a recording medium that holds a program for a certain period of time, such as a volatile memory (RAM) inside a computer system that becomes a server or client when a program is transmitted over a network. The program may also be a recording medium for implementing part of the above-mentioned functions. Furthermore, the program may be a so-called difference file (difference program) that can realize the above-mentioned functions in combination with a program already recorded in the computer system.

In addition, some or all of the above-mentioned functions may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each of the above-mentioned functions may be individually processed, or some or all of the functions may be integrated into a processor. The integrated circuit method is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. Furthermore, if an integrated circuit technology that can replace LSI appears due to advances in semiconductor technology, an integrated circuit based on that technology may be used.

1...Building frame estimation device, 11...Input unit, 12...Display unit, 13...Memory unit, 14...Control unit, 131...Learning data memory unit, 132...Model memory unit, 133...Input information memory unit, 134...Inner product information memory unit, 135...Building frame memory unit, 141...Model construction unit, 142...Data acquisition unit, 143...Inner product generation unit, 144...Building frame estimation unit, 145...Display control unit

Claims (7)

a model storage unit that stores a building frame estimation model for estimating a building frame of the building from a first joint number vector, which is a vector whose elements are the numbers of each type of joint connecting columns and beams of the building in a first direction parallel to a horizontal plane of the building, and a second joint number vector, which is a vector whose elements are the numbers of each type of joint in a second direction parallel to the horizontal plane and different from the first direction, a frame volume of the building , and basic design quantities taken into consideration in structural design ;
a building frame estimation unit that estimates a building frame of the building from the building frame estimation model stored in the model storage unit, the joint number vector inner product value, a frame volume of the building , and the basic design quantities ,
The building frame estimation model is constructed based on learning data having a plurality of sets of data including at least the joint number vector inner product value, the frame volume, the basic design quantities, and the building frame of the building.
A building frame estimation device characterized by : The junction number vector inner product value includes an inner product value of a top floor of the building, an inner product value of an intermediate floor of the building, and an inner product value of a bottom floor of the building;
The building frame estimation device according to claim 1, characterized in that the building frame estimation unit estimates the building frame of the building based on the inner product value of the top floor, the inner product value of the intermediate floor, and the inner product value of the bottom floor, the frame volume, and the basic design quantities . the building frame estimation model is a model for classifying the building frame into a plurality of groups by clustering based on the frame volume and the joint number vector inner product value, and outputting group identification information indicating the classified groups from the frame volume , the basic design quantities, and the joint number vector inner product value ;
The building frame estimation device according to claim 1 or claim 2, characterized in that the building frame estimation unit acquires the group identification information based on the building frame estimation model, and estimates the building frame belonging to the group corresponding to the acquired group identification information as the building frame of the building. a data acquisition unit that acquires the frame volume, the area of a reference floor of the building, the span length in the first direction, the span length in the second direction, and a floor height indicating a height of a floor of the building;
an inner product generating unit that generates the inner product value of the junction number vector based on the area of the reference floor, the span length in the first direction, the span length in the second direction, and the floor height acquired by the data acquiring unit;
The building frame estimation device according to claim 1 , further comprising: an output control unit that outputs, to an output unit, output information indicating the building frame of the building estimated by the building frame estimation unit. The types of the joints include a first joint that connects a column and a beam at a corner of the building, a second joint that connects a column and a beam on the outer periphery of the building, a third joint that connects a beam of the building to a column at either the top or bottom of the beam, and a fourth joint that connects the column and the beam of the building in a cross shape;
The building frame estimation device according to any one of claims 1 to 4, characterized in that the first joint number vector and the second joint number vector include the respective numbers of the first joints, the second joints, the third joints, and the fourth joints as vector elements. The building frame estimation device according to claim 1 , further comprising: a model construction unit that constructs the building frame estimation model based on the learning data. a joint number vector inner product value that is an inner product value of a first joint number vector, the first joint number vector being a vector whose elements are the numbers of each type of joint connecting columns and beams of the building in a first direction parallel to a horizontal plane of the building, and the first joint number vector, the first joint number vector being a vector whose elements are the numbers of each type of joint in a second direction parallel to the horizontal plane and different from the first direction, a model storage unit that stores a building frame estimation model for estimating a building frame of the building from a frame volume of the building and basic design quantities taken into consideration in structural design; and a building frame estimation unit,
the building frame estimation unit includes a building frame estimation step of estimating a building frame of the building from the building frame estimation model stored in the model storage unit, the joint number vector inner product value, a frame volume of the building , and the basic design quantities ,
The building frame estimation model is constructed based on learning data having a plurality of sets of data including at least the joint number vector inner product value, the frame volume, the basic design quantities, and the building frame of the building.
A building frame estimation method comprising :

Images