Architectural drawing design program and apparatus therefor

Abstract

To provide an architectural drawing design program and an apparatus therefor that can efficiently design architectural drawings satisfying a client's request.SOLUTION: An architectural drawing design program that designs architectural drawing data for a client causes a computer to realize: a data management function that receives the input of client feature data indicating features of the client with respect to a plurality of direct or indirect architectural elements and manages the data; a design function that, on the basis of the client feature data received from the data management function, utilizes a learned neural network that has learned interrelationships among respective architectural elements in client feature data related to past clients and architectural drawing data, to design architectural drawing data related to a building that satisfies the client's request; and an output control function that performs output control with the architectural drawing data as data learned by the learned neural network.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an architectural drawing design program and an apparatus therefor, and more particularly to an architectural drawing design program and an apparatus therefor that enable the design of a building that reflects the client's intentions with high precision. [Background technology]

[0002] For custom-built homes, construction companies and home builders hold interviews with the client and, based on the results, design architectural drawings.

[0003] However, the resulting architectural drawings, and in turn the building itself, do not necessarily meet the client's needs, and in many cases the client ends up compromising and living in the custom-built home, or ordering a new custom-built home in the future.

[0004] From this point of view, systematization of custom-built homes is being considered to enable efficient design of custom-built homes.

[0005] For example, conventional architectural drawing and design programs use pre-created virtual design specifications to virtually design buildings without a client, and then publish the resulting virtual design building on a network, allowing consumers to compare multiple virtual design buildings and, through consultations with the virtual designer, build a house that meets the wishes of the average consumer (see Patent Document 1).

[0006] In addition, for example, there is a housing design system that has a user terminal that allows a client to view architectural drawings of a house, and a contractor terminal that sends and receives data related to the architectural drawings to and from the user terminal, and the user terminal has a display unit that displays 3D design architectural drawings of the house, an editing unit that edits the architectural drawings displayed by the display unit based on operations by the client, and a transmission unit that transmits edited data including the edited content edited by the editing unit to the contractor terminal (see Patent Document 2).

[0007] For example, some conventional architectural drawing design programs use the artificial intelligence GAN (Generative Adversarial Networks) to automatically create a house floor plan based on the shape of the land (see Non-Patent Document 1). [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent Publication No. 2021-105815 [Patent Document 2] Japanese Patent Application Publication No. 2020-86809 [Non-patent literature]

[0009] [Non-Patent Document 1] https: / / www.affrc.maff.go.jp / docs / public_offering / agri_food2016 / 25055c.html Summary of the Invention [Problem to be solved by the invention]

[0010] However, when building a custom home, in addition to personal economic factors such as the client's budget, hobbies, and preferences, physical factors such as the land to be used, and legal factors such as legal restrictions on building height, are all intricately intertwined in a multifaceted manner.

[0011] From this point of view, for example, as in Patent Document 1, there are some systems in which a client compares multiple virtual design buildings that are published on the network, but there is no guarantee that the published virtual design buildings will include one that satisfies the client's requirements. While the client needs time and effort to find a virtual design building that meets their requirements, there is no guarantee that they will find one that meets their requirements.

[0012] In addition, for example, as in Patent Document 2, there are systems in which a client can view architectural drawings of a house in three dimensions on a user terminal, but the architectural drawings do not necessarily satisfy the client's requirements. As in the above, a client needs time and effort to find architectural drawings that satisfy their requirements, but they are not necessarily able to find those that satisfy their requirements.

[0013] Furthermore, for example, as in Non-Patent Document 1, there are some systems that use artificial intelligence, GAN (Generative Adversarial Network), to calculate an expected floor plan from the shape of the land, but this mechanically generates a floor plan from the shape of the land and does not reflect the client's intentions. As with the above, it takes time and effort for the client to find a floor plan that meets their requirements, but they are not necessarily able to find one that meets their requirements.

[0014] Although the systemization of custom-built homes is being considered, in reality it still takes time and effort to satisfy the diverse needs of clients.

[0015] The present invention has been made to solve the above-mentioned problems, and has as its object to provide an architectural drawing design program and an apparatus therefor that can efficiently design architectural drawings that satisfy the client's requests. [Means for solving the problem]

[0016] The architectural drawing design program disclosed in this application is an architectural drawing design program for designing architectural drawing data for a client, and has the following functions implemented in a computer: a data management function for accepting and managing input of client characteristic data that indicates the client's characteristics for multiple direct or indirect architectural elements; a design function for designing architectural drawing data for a building that meets the client's requirements based on the client characteristic data accepted from the data management function, using a trained neural network that has learned the interrelationships between the architectural elements of the client characteristic data and architectural drawing data for past clients; and an output control function for controlling the output of the architectural drawing data as data learned in the trained neural network.

[0017] In this way, the computer is provided with a data management function that accepts and manages input of client characteristic data that indicates the client's characteristics for multiple direct or indirect architectural elements, a design function that uses a trained neural network that has learned the interrelationships between the architectural elements of the client characteristic data and architectural drawing data for past clients based on the client characteristic data received from the data management function to design architectural drawing data for a building that meets the client's requirements, and an output control function that controls the output of the architectural drawing data as data learned in the trained neural network.Therefore, when client characteristic data is input, the trained neural network accumulated in the past is used to design architectural drawing data for a building that meets the client's requirements, making it possible to efficiently and accurately design architectural drawings that meet the client's needs.

[0018] In the architectural drawing design program disclosed in the present application, the data management function accepts and manages input of map data related to surveyed land, as necessary, and the design function includes a site determination function that identifies a site in the map data managed by the data management function as a building site based on at least the Building Standards Act and the City Planning Act. In this way, since the data management function accepts and manages input of map data related to surveyed land, and the design function includes a site determination function that identifies a site in the map data managed by the data management function as a building site based on at least the Building Standards Act and the City Planning Act, the design is based on a site where construction is actually possible, thereby improving the accuracy of the design and enabling the design of highly realistic architectural drawings that meet the client's needs.

[0019] In the architectural drawing design program disclosed in the present application, the data management function verifies the validity of the content of each architectural element of the input client characteristic data as necessary. In this way, the data management function verifies the validity of the content of each architectural element of the input client characteristic data, so that the input client characteristic data is judged to be consistent information, which improves compatibility with various conditions related to actual architecture, making it possible to design architectural drawings that more realistically meet the client's requests.

[0020] In the architectural drawing design program disclosed in the present application, the data management function determines the priority of each architectural element based on the interrelationships between the architectural elements in the input client characteristic data, as needed. In this way, the data management function determines the priority of each architectural element based on the interrelationships between the architectural elements in the input client characteristic data, which allows for the design of more flexible and realistic architectural drawing data in accordance with the priorities of individual clients, thereby enabling the design of more realistic architectural drawings that meet the client's needs.

[0021] In the architectural drawing design program disclosed in the present application, the design function optimizes the architectural drawing data generated by the trained neural network as needed to satisfy content validity. In this way, the design function optimizes the architectural drawing data generated by the trained neural network to satisfy content validity, resulting in the design of more realistic and valid architectural drawing data, and making it possible to design more realistic architectural drawings that satisfy the client's requests.

[0022] In the architectural drawing design program disclosed in the present application, the design function may include, as necessary, an external element determination function for determining external elements including the client's budget, a volume element determination function for determining volume elements including the site area of ​​the client's building based on the external elements, a zoning element determination function for determining zoning elements including plots related to the client's building based on the volume elements, and a planning element determination function for determining planning elements including the floor plan of the client's building based on the zoning elements. In this way, the design function includes an external element determination function that determines external elements including the client's budget, a volume element determination function that determines volume elements including the site area of ​​the client's building based on the external elements, a zoning element determination function that determines zoning elements including the plots related to the client's building based on the volume elements, and a planning element determination function that determines planning elements including the floor plan of the client's building based on the zoning elements, so that architectural drawing data for a building that meets the client's requirements can be designed in one step, from the land to the room layout, making it possible to efficiently design architectural drawings that meet the client's requests with a high degree of accuracy.

[0023] In the architectural drawing design program disclosed herein, the design function optionally includes a planning consistency determination function that determines the consistency of the planning elements determined by the planning element determination function as a custom-built home, and if the planning consistency determination function determines that consistency is not met, the zoning element determination function and the planning element determination function are executed again to determine new planning elements. In this way, the design function includes a planning consistency determination function that determines the consistency of the planning elements determined by the planning element determination function as a custom-built home, and if the planning consistency determination function determines that consistency is not met, the zoning element determination function and the planning element determination function are executed again to determine new planning elements. Therefore, the planning elements are repeatedly determined until consistency as a custom-built home is achieved, making it possible to meet the client's needs with content that takes advantage of the unique characteristics of custom-built homes, which are different from pre-built homes and the like, and to design architectural drawings that provide greater customer satisfaction.

[0024] In the architectural drawing design program disclosed herein, the design function optionally includes a zoning consistency determination function for determining whether the zoning elements determined by the zoning element determination function are consistent with the architecture of a custom-built home, and if the zoning consistency determination function determines that consistency is not met, the zoning element determination function is executed again to determine new zoning elements. In this way, the design function includes a zoning consistency determination function for determining whether the zoning elements determined by the zoning element determination function are consistent with the architecture of a custom-built home, and if the zoning consistency determination function determines that consistency is not met, the zoning element determination function is executed again to determine new zoning elements. Therefore, the zoning elements are repeatedly determined until a plot that is consistent with the architecture of a custom-built home is obtained, making it possible to meet the client's needs with content that takes advantage of the unique characteristics of custom-built homes, which are different from prefabricated homes and the like, and to design architectural drawings that provide greater customer satisfaction.

[0025] The architectural drawing design program disclosed in the present application causes a computer to, as needed, realize a virtual space construction function that constructs a virtual space based on architectural drawing data designed by the design function, and the output control function controls the output of the virtual space. In this way, by causing a computer to realize a virtual space construction function that constructs a virtual space based on architectural drawing data designed by the design function, and the output control function controls the output of the virtual space, it becomes possible to experience living in a building based on architectural drawing data designed by the design function in advance in the virtual space, allowing the client's requests to be concretely experienced in advance, thereby increasing the client's reliability in the architectural drawing data and further increasing customer satisfaction.

[0026] In the architectural drawing design program disclosed in the present application, the client characteristic data is formed as a distribution element consisting of each of the architectural elements, as needed. In this way, the client characteristic data is formed as a distribution element consisting of each of the architectural elements, so that the client's characteristics can be immediately and accurately grasped from the distribution element, thereby increasing the client's reliability in the architectural drawing data according to the client's characteristics and further increasing customer satisfaction.

[0027] In the architectural drawing design program disclosed in the present application, if necessary, when there are multiple clients, the design function uses a trained neural network to design architectural drawing data for a building that meets the requirements of the multiple clients, based on the multiple distribution elements corresponding to the multiple client characteristic data. In this way, when there are multiple clients, the design function uses a trained neural network to design architectural drawing data for a building that meets the requirements of the multiple clients, based on the multiple distribution elements corresponding to the multiple client characteristic data. Therefore, even when there are multiple clients, it does not take a lot of time and effort to consolidate the many different requirements of each client into one, and architectural drawing data that consolidates the individual requirements of each client can be efficiently designed in a shorter time, and customer satisfaction can be further increased through time-efficient designs.

[0028] The architectural drawing design device disclosed in this application is an architectural drawing design device that designs architectural drawing data for a client, and is equipped with: a data management means that accepts and manages input of client characteristic data that indicates the client's characteristics for multiple direct or indirect architectural elements; a design means that, based on the client characteristic data accepted from the data management means, designs architectural drawing data for a building that meets the client's requirements by utilizing a trained neural network that has learned the interrelationships between the architectural elements of the client characteristic data and architectural drawing data for past clients; and an output control means that controls the output of the architectural drawing data as data learned by the trained neural network.

[0029] In this way, the architectural drawing design device that designs architectural drawing data for a client comprises: a data management means that accepts and manages input of client characteristic data that indicates the client's characteristics for multiple direct or indirect architectural elements; a design means that, based on the client characteristic data accepted from the data management means, uses a trained neural network that has learned the interrelationships between the architectural elements of the client characteristic data and architectural drawing data for past clients to design architectural drawing data for a building that meets the client's requirements; and an output control means that controls the output of the architectural drawing data as data learned in the trained neural network.Therefore, when client characteristic data is input, the trained neural network accumulated in the past is used to design architectural drawing data for a building that meets the client's requirements, making it possible to efficiently and accurately design architectural drawings that meet the client's needs.

[0030] In the architectural drawing design device disclosed in the present application, the data management means accepts and manages input of map data related to surveyed land as needed, and the design means includes a site determination means for calculating the extent of a buildable site from the map data managed by the data management means, based on at least the Building Standards Act and the City Planning Act. In this way, since the data management means accepts and manages input of map data related to surveyed land and the design means includes a site determination means for calculating the extent of a buildable site from the map data managed by the data management means, based on at least the Building Standards Act and the City Planning Act, the design is based on a realistically buildable site, which improves the accuracy of the design and makes it possible to design architectural drawings that more realistically meet the client's needs.

[0031] In the architectural drawing design device disclosed in the present application, the data management means verifies the validity of the content of each architectural element of the input client characteristic data as necessary. In this way, since the data management means verifies the validity of the content of each architectural element of the input client characteristic data, the input client characteristic data is judged to be consistent information, and compatibility with various conditions related to actual architecture is improved, making it possible to design architectural drawings that more realistically meet the client's requests.

[0032] In the architectural drawing design device disclosed in the present application, the data management means determines the priority of each architectural element based on the interrelationships between the architectural elements in the input client characteristic data, as needed. In this way, the data management means determines the priority of each architectural element based on the interrelationships between the architectural elements in the input client characteristic data, so that flexible and highly realistic architectural drawing data can be designed in accordance with the priorities of individual clients, and more realistic architectural drawings can be designed to meet the client's needs.

[0033] In the architectural drawing design device disclosed in the present application, the design means adjusts the architectural drawing data generated by the trained neural network as needed to satisfy content validity. In this way, the design means adjusts the architectural drawing data generated by the trained neural network to satisfy content validity, so that more realistic and valid architectural drawing data can be designed, and more realistic architectural drawings can be designed to satisfy the client's requests.

[0034] In the architectural drawing design device disclosed in the present application, the design means includes, as necessary, an external element determination means for determining external elements including the client's budget, a volume element determination means for determining a volume element including a site area related to the client's building based on the external elements, a zoning element determination means for determining a zoning element including a section related to the client's building based on the volume elements, and a planning element determination means for determining a planning element including a floor plan related to the client's building based on the zoning elements. In this way, the design means includes an external element determination means for determining external elements including the client's budget, a volume element determination means for determining a volume element including the site area of ​​the client's building based on the external elements, a zoning element determination means for determining a zoning element including a section of the client's building based on the volume elements, and a planning element determination means for determining a planning element including the floor plan of the client's building based on the zoning elements.As a result, architectural drawing data for a building that meets the client's requirements can be designed in one step, from the land to the room layout, and architectural drawings that meet the client's requests can be designed efficiently and accurately.

[0035] In the architectural drawing design device disclosed in the present application, the design means may, as necessary, include a planning consistency determination means for determining the consistency of the planning elements determined by the planning element determination means as a building for a custom-built home, and if the planning consistency determination means determines that consistency is not met, the zoning element determination means and the planning element determination means are executed again to determine new planning elements. In this way, the design means may include a planning consistency determination means for determining the consistency of the planning elements determined by the planning element determination means as a building for a custom-built home, and if the planning consistency determination means determines that consistency is not met, the zoning element determination means and the planning element determination means are executed again to determine new planning elements. Therefore, the planning elements are repeatedly determined until consistency as a floor plan for a custom-built home is achieved, making it possible to meet the client's needs with content that makes use of the unique characteristics of custom-built homes that differ from those of prefabricated homes, etc., and to design architectural drawings that provide greater customer satisfaction.

[0036] In the architectural drawing design device disclosed in the present application, the design means is, as necessary, equipped with a zoning consistency determination means for determining the consistency of the zoning elements determined by the zoning element determination means as a building for a custom-built home, and if the zoning consistency determination means determines that consistency is not met, the zoning element determination means is executed again to determine new zoning elements. In this way, the design means is equipped with a zoning consistency determination means for determining the consistency of the zoning elements determined by the zoning element determination means as a building for a custom-built home, and if the zoning consistency determination means determines that consistency is not met, the zoning element determination means is executed again to determine new zoning elements. Therefore, the zoning elements are repeatedly determined until consistency as a plot for a custom-built home is achieved, making it possible to meet the client's needs with content that makes use of the unique characteristics of custom-built homes that differ from those of prefabricated homes, etc., and to design architectural drawings that provide greater customer satisfaction.

[0037] The architectural drawing design device disclosed in the present application comprises a virtual space construction means for constructing a virtual space based on architectural drawing data designed by the design means as needed, and the output control means controls the output of the virtual space. As described above, the device comprises a virtual space construction means for constructing a virtual space based on architectural drawing data designed by the design means, and the output control means controls the output of the virtual space. This allows a user to experience living in a building based on the architectural drawing data designed by the design means in advance in the virtual space, allowing the user to concretely experience the client's wishes in advance, thereby increasing the client's reliability in the architectural drawing data and further increasing customer satisfaction.

[0038] In the architectural drawing design device disclosed in the present application, the client characteristic data is formed as a distribution element consisting of each of the architectural elements, as needed. In this way, the client characteristic data is formed as a distribution element consisting of each of the architectural elements, so that the client's characteristics can be immediately and accurately grasped from the distribution element, thereby increasing the client's reliability in the architectural drawing data according to the client's characteristics and further increasing customer satisfaction.

[0039] In the architectural drawing design device disclosed in the present application, if necessary, when there are multiple clients, the design means uses a trained neural network to design architectural drawing data for a building that meets the requirements of the multiple clients, based on the multiple distribution elements corresponding to the multiple client characteristic data. In this way, when there are multiple clients, the design means uses a trained neural network to design architectural drawing data for a building that meets the requirements of the multiple clients, based on the multiple distribution elements corresponding to the multiple client characteristic data. Therefore, even when there are multiple clients, it does not take a lot of time and effort to consolidate the many different requirements of each client into one, and architectural drawing data that consolidates the individual requirements of each client can be efficiently designed in a shorter time, and customer satisfaction can be further increased through time-efficient designs. [Brief explanation of the drawings]

[0040] [Figure 1] 1 is a diagram showing a system configuration of an architectural drawing design device according to a first embodiment. [Figure 2] 1 is a functional block diagram showing a configuration of an architectural drawing design device according to a first embodiment. [Figure 3] FIG. 2 is an explanatory diagram showing the processing of a design unit of the architectural drawing design device according to the first embodiment. [Figure 4] 3 is an explanatory diagram showing an example of architectural drawing data output by an output control unit of the architectural drawing design device according to the first embodiment; FIG. [Figure 5] 3 is a flowchart showing the operation of the architectural drawing design device according to the first embodiment. [Figure 6] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a second embodiment. [Figure 7] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a second embodiment. [Figure 8] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a third embodiment. [Figure 9] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a fourth embodiment. [Figure 10] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a fifth embodiment. [Figure 11] FIG. 10 is a diagram showing a system configuration of an architectural drawing design device according to a fifth embodiment. [Figure 12] 13 is a flowchart showing the operation of an architectural drawing design device according to a sixth embodiment. [Figure 13] FIG. 2 is a diagram showing client characteristic data input to the architectural drawing design device according to the first embodiment of the present invention. [Figure 14] FIG. 2 is a diagram showing a design result of the architectural drawing design device according to the first embodiment of the present invention. [Figure 15] FIG. 2 is a diagram showing a design result of the architectural drawing design device according to the first embodiment of the present invention. [Figure 16] FIG. 10 is a diagram showing client characteristic data input to an architectural drawing design device according to a second embodiment of the present invention. [Figure 17] FIG. 10 is a diagram showing a design result of the architectural drawing design device according to the second embodiment of the present invention. [Figure 18] FIG. 10 is a diagram showing a design result of the architectural drawing design device according to the second embodiment of the present invention. [Figure 19] FIG. 10 is a diagram showing client characteristic data input to an architectural drawing design device according to a third embodiment of the present invention. [Figure 20] FIG. 10 is a diagram showing a design result of the architectural drawing design device according to the third embodiment of the present invention. [Figure 21] FIG. 10 is a diagram showing a design result of the architectural drawing design device according to the third embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0041] Hereinafter, embodiments of the present invention will be described. Similar elements throughout the embodiments are designated by the same reference numerals.

[0042] (First embodiment of the present invention) An architectural drawing design device 10 according to this embodiment will be described with reference to the above-mentioned Figs. 1 to 4. The architectural drawing design device 10 according to this embodiment designs architectural drawings in accordance with the wishes of a client, for example, when the client consults a housing manufacturer about a custom-built house. In this embodiment, an architectural drawing design device 10 will be described that can design architectural drawings for a custom-built house in accordance with the wishes of a client (owner) when the client consults a custom-built house.

[0043] Fig. 1 shows the system configuration of an architectural drawing design device 10 according to this embodiment. As shown in Fig. 1, the architectural drawing design device 10, which designs architectural drawing data 200 for a client, reads client characteristic data 100 indicating the characteristics of the client, and outputs architectural drawing data 200 that meets the client's wishes.

[0044] As shown in Figure 2, the architectural drawing design device 10 according to this embodiment is configured to include a data management unit 11 that receives and manages input of client characteristic data 100 that indicates the client's characteristics regarding multiple direct or indirect architectural elements, a design unit 12 that uses a trained neural network to design architectural drawing data 200 for a building that meets the client's requirements based on the client characteristic data 100 received from the data management unit 11, and an output control unit 13 that controls the output of this architectural drawing data 200.

[0045] The output control unit 13 serves as a user interface, displaying the designed architectural drawing data 200 to the user on an external display unit such as a display. This display is not particularly limited, but examples include a CRT monitor, a liquid crystal monitor, and an organic EL monitor. In addition to output to this screen, the output control unit 13 can also control output as data. For example, the output control unit 13 also controls output, such as sending result data as a reply in response to a request from a client PC.

[0046] Here, there are multiple clients who have previously designed other architectural drawing data 200 using the program according to this embodiment. The trained neural network has learned the interrelationships between the architectural elements of the client characteristic data 100 related to the past clients and the architectural drawing data 200.

[0047] The client characteristic data 100 indicates the client's characteristics regarding multiple direct or indirect building elements, such as budget, annual income, hobbies (e.g., outdoors, watching sports, watching movies, reading, cooking, DIY, gardening, no hobbies), family composition (adults, children), priorities (time spent together, time spent as a couple, family dinners, hobbies in individual spaces, time spent by family members, hosting parties), number of rooms, and specific requirements, which may be numerical or non-numerical information.

[0048] The generated architectural drawing data 200 is not particularly limited, but may be represented visually using commercially available image rendering software, or may be represented numerically as binary code equivalent to an image.

[0049] In Figure 2, a neural network has a layered structure of interconnected nodes and neurons that mimics the human brain, and represents a type of machine learning process such as deep learning.

[0050] The neural network of this embodiment has a structure including an input layer, multiple intermediate layers, and an output layer, and can also be a multi-layered neural network. A neural network based on deep learning, a type of machine learning, can also be used. Each neuron in the input layer is provided with input items that make up the client characteristic data 100, and receives the input items entered by the client as, for example, numerical values. Each neuron in the output layer outputs the content of each item in the output items as, for example, a numerical value.

[0051] Preferably, the data management unit 11 judges the validity of the content of each architectural element in the input client characteristic data 100. For example, it judges whether the data conforms to current building laws. It performs a consistency check, a process known as a consistency filter. This results in the client characteristic data 100 being revised to be more realistic, improving the accuracy of the input information to the trained neural network and enabling the design of high-quality architectural drawing data 200. This validity check can be performed by the check processing unit 11a, as shown in Figure 2.

[0052] The check processing unit 11a judges the validity of the content of each building element of the input client characteristic data 100 based on information such as laws and regulations related to building design matters and appropriate building coverage ratios, and performs correction calculations to optimize the content. This correction calculation process may be automatic processing using an algorithm, or may involve manual input by an operator.

[0053] Preferably, the data management unit 11 determines the priority of each architectural element based on the interrelationships between the architectural elements in the input client characteristic data 100. For example, if the client places importance on hobbies as architectural elements, it will give higher priority to items related to those hobbies, such as the size of the garden if the client's hobby is gardening, or the size of the kitchen if the client's hobby is cooking.

[0054] In addition to such a unique algorithm-based method, it is also possible to use information obtained when a sales representative interviews a client. For example, the time it takes to answer each question, the conversation time, and the frequency of each keyword during the client interview can be used. Based on this priority, the client characteristic data 100 is modified to be closer to the client's wishes, and the input information to the trained neural network is made more accurate, enabling the design of high-quality architectural drawing data 200.

[0055] 2, the architectural drawing design program according to this embodiment is realized by causing the CPU, which is the calculation unit of a computer, to function as each processing unit, namely, the data management unit 11, the check processing unit 11a, the design unit 12, and the output control unit 13. In particular, the data input / output functions of the data management unit 11 and the output control unit 13 are realized by causing them to function as input / output interface units.

[0056] The processing content of each processing unit will be described in detail below. Figure 3 is a diagram showing the processing of the design unit 12 of the architectural drawing design device 10 according to this embodiment. As shown in Figure 3, numerical values ​​representing, for example, budget, hobbies, and annual income (hereinafter referred to as input values) are input to the data management unit 11 as client characteristic data 100 of the client. For example, if the budget is 10 million yen, the hobby is gardening, and the annual income is 6 million yen, a numerical value that uniquely represents these is used as this input value. For example, this can be freely designed by program design, such as "100002600" which combines the numerical value "1000" representing a budget of 10 million yen, the numerical value "02" representing the hobby of gardening, and the numerical value "600" representing an annual income of 6 million yen.

[0057] The design unit 12 uses this input value "100002600" to design architectural drawing data 200 that meets the client's (owner's) requirements using a trained neural network that has accumulated correlations between the client characteristic data 100 and the architectural drawing data 200 that have been accumulated in the past.

[0058] From these input values, the trained neural network generates architectural drawing data 200. This architectural drawing data 200 is output, for example, as binary data representing architectural drawings.

[0059] As an example of the computation of the trained neural network, if there is one past example of architectural drawing data 200 generated by inputting the same value as this input value, that example is adopted. If there are multiple past examples of architectural drawing data 200 generated by inputting the same value as this input value, for example, a process is performed in which commonalities are retained and differences are averaged among the binary data representing the architectural drawings. If there are no past examples of architectural drawing data 200 generated by inputting the same value as this input value, for example, a process is performed in which multiple examples of architectural drawing data 200 similar to this input value are collected, and commonalities are retained and differences among the binary data representing these architectural drawings are averaged. For example, if the binary data representing the architectural drawings consists of three values, 001100, 001101, and 001101, the first five bytes that are common to all three are retained, and the remaining differences are averaged, and the architectural drawing data 200 represented by 001101 is adopted.

[0060] Note that the input value "100002600" is merely an example, and other expressions are possible. Also, the input value is not limited to the budget, hobbies, and annual income as in the above example, but can include various elements of the client.

[0061] 4 shows an example of architectural drawing data 200 output by the output control unit 13 of the architectural drawing design device 10 according to this embodiment. The architectural drawing design device 10 may generate architectural drawing data 200 in the form of visual drawing data or numerical data, which can be output by the output control unit 13. The generated architectural drawing data 200 may be displayed on a display screen, printed on paper, or output as a data file to an electronic medium such as an external memory.

[0062] As shown in Figure 4, the land parcel and the floor plan of the custom-built house that meets the client's (owner's) requests can be visually recognized based on the generated architectural drawing data 200. That is, the client's (owner's) requests may result in a floor plan such as that shown in Figure 4(a) being designed, or the client's (owner's) requests may result in a floor plan such as that shown in Figure 4(b).

[0063] If there are multiple pieces of architectural drawing data 200 generated by the design unit 12, the output control unit 13 can also list and display multiple pieces of data at the same time. This allows the client (owner) to narrow down and select their preferred architectural drawings, providing a highly convenient service.

[0064] Next, the operation of the architectural drawing design device 10 according to this embodiment will be described with reference to the flowchart of Fig. 5. First, the data management unit 11 reads the client characteristic data 100, which is input information (S1). This client characteristic data 100 may be input directly by the client (owner) into an input terminal, or a person in charge at a custom home builder may hold a hearing with the client (owner) and input the data into the input terminal based on the results of the hearing.

[0065] Next, the design unit 12 uses the trained neural network to design architectural drawing data 200 for a building that meets the client's requirements (S2). As described above, by using this trained neural network, highly accurate architectural drawing data 200 that meets the client's (client's) needs is designed. At this point, the client characteristic data 100 and architectural drawing data 200 of the client (client) are learned by the neural network as past design results.

[0066] More preferably, the design unit 12 optimizes the architectural drawing data 200 generated by the trained neural network to satisfy content validity. That is, if the generated architectural drawing data 200 does not comply with the Building Act, it is determined that the content is invalid, and the trained neural network is caused to recalculate again until the content validity is satisfied. The trained neural network weights the generated architectural drawing data 200 based on the presence or absence of validity and trains it, thereby improving the accuracy of subsequent designs. From the next design onward, designs are performed in descending order of weighting, further improving the accuracy of the design. That is, this optimization results in the design of more realistic and valid architectural drawing data 200, enabling the design of more realistic architectural drawings that meet the client's needs.

[0067] Next, the output control unit 13 controls the display of the designed architectural drawing data 200 (S3), thereby completing the processing for the client (owner).

[0068] Furthermore, regarding the display of this output control unit 13, for example, a virtual space construction unit that constructs a virtual space based on the architectural drawing data 200 designed by the design unit 12 may be provided, and the output control unit 13 may control the output of this virtual space. As the virtual space, a metaverse or the like may be used.

[0069] With this configuration, a building based on the architectural drawing data 200 designed by this design means can be experienced in advance in a virtual space, allowing the client's requests to be concretely experienced in advance, increasing the client's reliability in the architectural drawing data 200 and further increasing customer satisfaction.

[0070] In this way, when client characteristic data 100 is input into the architectural drawing design device 10 according to this embodiment, architectural drawing data 200 relating to a building that meets the client's requirements is designed using a trained neural network that has been accumulated in the past, and architectural drawings that meet the client's needs can be designed efficiently and accurately.

[0071] Furthermore, the architectural drawing design device 10 according to the present embodiment is not particularly limited as long as it is a device in which the architectural drawing design program according to the present embodiment is installed, and includes, for example, a device installed on a PC client terminal or a cloud server terminal such as IaaS, PaaS, or SaaS. In this way, once installed, the architectural drawing design program according to the present embodiment can run on a PC terminal or on the cloud. Furthermore, output information can be output to a screen display on the terminal, or can be output as output data in response to a request to a server.

[0072] (Second embodiment of the present invention) The architectural drawing design device 10 according to this embodiment will be described with reference to Figures 6 and 7. In the architectural drawing design device 10 according to this embodiment, the design unit 12 performs more stepwise subdivision processing.

[0073] That is, similar to the first embodiment described above, it is equipped with a data management unit 11, a design unit 12, and an output control unit 13, and furthermore, this design unit 12 is equipped with an external element determination unit 12a that determines external elements including the budget of this other client, a volume element determination unit 12b that determines volume elements including the site area of ​​this other client's building based on these external elements, a zoning element determination unit 12c that determines zoning elements including the plots related to this other client's building based on these volume elements, and a planning element determination unit 12d that determines planning elements including the floor plan of this other client's building based on these zoning elements.

[0074] The processing content of each processing unit will be described in detail below. Fig. 6 is a diagram showing the processing of the design unit 12 of the architectural drawing design device 10 according to this embodiment.

[0075] As shown in FIG. 6, numerical values ​​representing, for example, budget, hobbies, and annual income (hereinafter referred to as input values) are input to the data management unit 11 as client characteristic data 100 of the client (client).

[0076] (External element judgment unit 12a) The external element determination unit 12a determines the external elements based on the trained neural network on the basis of external related information, such as input information on budget, age, family structure, Feng Shui orientation, and applicable laws, from the input client characteristic data 100. For example, the budget may be the client's upper limit, and the Feng Shui orientation may be a level of trust in Feng Shui, etc.

[0077] The external elements are elements of the architectural drawing data 200, which is the final output, that correspond to the prerequisites for constructing a building, such as the land area, land orientation, land height, etc. Information on these external elements may be visually expressed as drawing data using image rendering software, or may be numerically expressed as binary code equivalent to an image.

[0078] Regarding the processing content, numerical values ​​representing, for example, land area, land orientation, land height, etc. (hereinafter referred to as input values) are input from the data management unit 11 as the client's client characteristic data 100. For example, if the land area is 100 tsubo, the land orientation is south, and the land height is 5 m, a numerical value that uniquely represents these is used as this input value. For example, this can be freely designed by program design, such as "10002005," which combines the numerical value "100" representing the land area of ​​100 tsubo, the numerical value "02" representing the land orientation facing south, and the numerical value "005" representing the land height of 5 m.

[0079] The design department 12 uses this input value "10002005" to design architectural drawing data 200 that meets the client's (owner's) requirements using a trained neural network that has accumulated correlations between the client characteristic data 100 and the architectural drawing data 200 that have been accumulated in the past.

[0080] The trained neural network generates an extrinsic element from the input value, and the extrinsic element is output as, for example, binary data representing the extrinsic element.

[0081] As an example of the computation of a trained neural network, if there is one past example of an external element generated by an input with the same value as this input, that example is adopted. If there are multiple past examples of external elements generated by inputs with the same value as this input, for example, a process is performed in which the commonalities are retained and differences are averaged among the binary data representing the external elements. If there are no past examples of external elements generated by inputs with the same value as this input, for example, a process is performed in which multiple examples of external elements similar to this input value are collected, and the commonalities are retained and differences are averaged among the binary data representing these architectural drawings. For example, if the binary data representing architectural drawings consists of three values, 001100, 001101, and 001101, the first five bytes that are common to all are retained, and the remaining differences are averaged to adopt the external element represented by 001101.

[0082] Note that this input value "10002005" is merely an example, and other expressions are possible. Furthermore, the input value is not limited to the land area, land orientation, and land height as in the above example, but can also include various client factors, such as budget, personal funds, borrowing status, desired repayment amount, desired building performance, and number of owned cars.

[0083] (Volume element determination unit 12b) The volume element determination unit 12b determines the volume elements including the budgets of other clients based on the input client characteristic data 100 and the external elements determined by the external element determination unit 12a, for example, based on volume-related information such as input information on family structure, budget, age, family structure, and feng shui orientation, and based on a trained neural network.

[0084] Volume elements are elements that correspond to the outer frame of the building to be constructed in the architectural drawing data 200, which is the final output, and, assuming the above external elements, correspond to, for example, the overall shape, upper and lower floors, orientation, height, etc. Information as these volume elements may be visually represented as drawing data using image rendering software, or may be numerically represented as binary code equivalent to an image.

[0085] Regarding the processing details, the volume element determination unit 12b receives input of external elements from the external element determination unit 12a, and also receives input of numerical values ​​representing, for example, family structure, budget, age, family structure, Feng Shui orientation, etc. (hereinafter referred to as input values) from the data management unit 11 as the client characteristic data 100 of the client. In the case of family structure, budget, age, family structure, and Feng Shui orientation, numerical values ​​that uniquely represent these are used as input values. For example, if the family structure is 5 people, the budget is 6 million yen, and the age is 30, the input values ​​can be freely designed by program design, such as "005600030," which combines the numerical value "005" representing the family structure of 5 people, the numerical value "600" representing the budget of 6 million yen, and the numerical value "030" representing the age of 30.

[0086] The volume element determination unit 12b of the design unit 12 determines from this input value "005600030" The correlation between the client characteristic data 100 and the architectural drawing data 200 that have been accumulated in the past is accumulated. Using a trained neural network, the volume that matches the client's (owner's) request is calculated. This external element is output as binary data indicating the external element, for example. Be encouraged.

[0087] As an example of the computation of a trained neural network, if there is one past example of a volume element generated by an input with the same value as the input value, that example is adopted. If there are multiple past examples of volume elements generated by an input with the same value as the input value, for example, a process is performed in which commonalities are retained and differences are averaged among the binary data representing the volume elements. If there are no past examples of volume elements generated by an input with the same value as the input value, for example, a process is performed in which multiple examples of volume elements similar to the input value are collected, and commonalities are retained and differences are averaged among the binary data representing these architectural drawings. For example, if the binary data representing architectural drawings consists of three values, 001100, 001101, and 001101, the first five bytes that are common to all are retained, and the remaining differences are averaged to adopt the volume element represented by 001101.

[0088] Note that this input value "005600030" is merely an example, and other expressions are possible. Also, the input value is not limited to the family structure, budget, and age as in the above example, but can include various client factors, such as the desired lot area.

[0089] (Zoning element determination unit 12c) The zoning element determination unit 12c determines zoning elements including sections related to the building based on the input client characteristic data 100 and the volume elements determined by the volume element determination unit 12b, and based on zoning-related information such as input information such as hobbies, family composition, budget, age, family composition, and feng shui orientation, and based on a trained neural network.

[0090] Zoning elements are elements that correspond to the inner frame of the building to be constructed in the architectural drawing data 200, which is the final output, and based on the volume elements described above, correspond to, for example, the building's internal divisions, internal layout, upper and lower floors, orientation, height, etc. Information on these zoning elements may be visually expressed as drawing data using image rendering software, or may be numerically expressed as binary code equivalent to an image.

[0091] Regarding the processing details, the zoning element determination unit 12c receives input of the volume elements from the volume element determination unit 12b, and also receives input of numerical values ​​representing the client's client characteristic data 100, such as family structure, budget, age, family structure, and Feng Shui orientation (hereinafter referred to as input values), from the data management unit 11. In the case of family structure, budget, age, family structure, and Feng Shui orientation, numerical values ​​that uniquely represent these are used as input values. For example, if the family structure is 5 people, the budget is 6 million yen, and the age is 30, the input value can be freely designed by program design, such as "005600030," which combines the numerical value "005" representing the family structure of 5 people, the numerical value "600" representing the budget of 6 million yen, and the numerical value "030" representing the age of 30.

[0092] The zoning element determination unit 12c of the design unit 12 determines the zoning elements that meet the client's (owner's) requirements from this input value "005600030" using a trained neural network that has accumulated correlations between the client characteristic data 100 and the architectural drawing data 200 that have been accumulated in the past. This external element is output, for example, as binary data indicating the external element.

[0093] As an example of the computation of a trained neural network, if there is one past example of a zoning element generated by an input with the same value as this input, that example is adopted. If there are multiple past examples of zoning elements generated by an input with the same value as this input, for example, a process is performed in which the commonalities are retained and differences are averaged among the binary data representing the zoning elements. If there are no past examples of zoning elements generated by an input with the same value as this input, for example, a process is performed in which multiple examples of zoning elements similar to this input value are collected, and the commonalities are retained and differences are averaged among the binary data representing these architectural drawings. For example, if the binary data representing architectural drawings consists of three values, 001100, 001101, and 001101, the first five bytes that are common to all three are retained, and the remaining differences are averaged to adopt the zoning element represented by 001101.

[0094] (Planning element determination unit 12d) The planning element determination unit 12d determines the planning elements based on the input client characteristic data 100 and the zoning elements determined by the zoning element determination unit 12c, and based on planning-related information such as input information such as hobbies, family structure, budget, age, family structure, and feng shui orientation, and on a trained neural network.

[0095] Planning elements are elements of the architectural drawing data 200, which is the final output, that correspond to the interior layout of the building to be constructed, and, assuming the above-mentioned zoning elements, correspond to, for example, the layout, upper and lower floors, orientation, height, etc. of the interior of the building. Information as these planning elements may be visually expressed as drawing data using image rendering software, or may be numerically expressed as binary code equivalent to an image.

[0096] Regarding the processing details, the planning element determination unit 12d receives zoning elements from the zoning element determination unit 12c, and also receives numerical values ​​representing the client's client characteristic data 100, such as family structure, budget, age, family structure, and Feng Shui orientation (hereinafter referred to as input values), from the data management unit 11. For the family structure, budget, age, family structure, and Feng Shui orientation, numerical values ​​that uniquely represent these are used as input values. For example, if the family structure is five people, the budget is 6 million yen, and the age is 30, the input value can be freely designed by program design, such as "005600030," which combines the numerical value "005" representing the family structure of five people, the numerical value "600" representing the budget of 6 million yen, and the numerical value "030" representing the age of 30.

[0097] The planning element determination unit 12d of the design unit 12 determines the planning element that matches the client's (owner's) request from this input value "005600030" using a trained neural network that stores correlations between the client characteristic data 100 and the architectural drawing data 200 that have been previously stored. This external element is output, for example, as binary data indicating the external element.

[0098] As an example of the computation of a trained neural network, if there is one past example of a planning element generated by an input with the same value as the input value, that example is adopted. If there are multiple past examples of a planning element generated by an input with the same value as the input value, for example, a process is performed in which the commonalities are retained and differences are averaged among the binary data representing the planning elements. If there are no past examples of a planning element generated by an input with the same value as the input value, for example, a process is performed in which multiple examples of planning elements similar to the input value are collected, and the commonalities are retained and differences are averaged among the binary data representing these architectural drawings. For example, if the binary data representing architectural drawings consists of three values, 001100, 001101, and 001101, the first five bytes that are common to all three are retained, and the remaining differences are averaged to adopt the planning element represented by 001101.

[0099] In this way, the architectural drawing data 200 shown in Fig. 4 above is finally output by the output control unit 13 of the architectural drawing design device 10 according to this embodiment. The output control unit 13 of the architectural drawing design device 10 can output the generated architectural drawing data 200 whether it is numerical data or visual drawing data. The generated architectural drawing data 200 may be displayed on a display screen, printed on paper media, or output as a data file to electronic media such as external memory.

[0100] More preferably, the system includes element input content preprocessing units (external element input content preprocessing unit 12A, volume element input content preprocessing unit 12C, zoning element input content preprocessing unit 12E, and zoning element input content preprocessing unit 12G) that preprocess input content to each of the above-mentioned judgment units to ensure that the content satisfies the requirements. As shown in FIG. 7 , each element input content preprocessing unit determines that the input content is invalid if, for example, it does not comply with the Building Code or does not have an appropriate building coverage ratio, and corrects the input content to ensure that the input content satisfies the requirements. This correction logic can be performed automatically using an algorithm. Alternatively, an operator can manually input the input content, for example, to reflect the latest legal changes.

[0101] More preferably, each of the judgment units in the design unit 12 optimizes the architectural drawing data 200 generated by the trained neural network or its equivalent elements so that the content satisfies the validity of the content. That is, as shown in Figure 7, if the generated architectural drawing data 200 does not comply with the Building Act, it is judged that the content is invalid, and the trained neural network is made to perform recalculation again until the content satisfies the validity of the content.

[0102] As shown in Figure 7, the external element determination unit 12a generates external elements using the trained neural network, or the information on the architectural drawing data 200 at this point, and the external element optimization unit 12B optimizes the information so that the content meets the validity criteria. This optimization may be performed manually, for example. Preferably, the priorities described in the first embodiment are used to lower the weighting or delete items with low priorities, causing the trained neural network to recalculate until the content meets the validity criteria.

[0103] 7, the volume element determination unit 12b generates volume elements using the trained neural network, or the information on the architectural drawing data 200 at this point in time. The volume element optimization unit 12D optimizes the volume elements so that the content meets the validity criteria. This optimization may be performed manually. Preferably, the priority ranking described in the first embodiment is used to reduce or delete items with low priority, causing the trained neural network to recalculate until the content meets the validity criteria.

[0104] 7, the zoning element optimization unit 12F optimizes the zoning elements generated by the zoning element determination unit 12c using the trained neural network or the information in the architectural drawing data 200 at this point so that the content meets the validity criteria. This optimization may be performed manually, for example. Preferably, the priority ranking described in the first embodiment is used to reduce or delete items with low priority, causing the trained neural network to recalculate until the content meets the validity criteria.

[0105] 7, the planning element optimization unit 12H optimizes the planning elements generated by the planning element determination unit 12d using the trained neural network or the information on the architectural drawing data 200 at this point so that the content meets the validity criteria. This optimization may be performed manually, for example. Preferably, the priority ranking described in the first embodiment is used to reduce or delete items with low priority, causing the trained neural network to recalculate until the content meets the validity criteria.

[0106] The trained neural network weights the generated architectural drawing data 200 based on the presence or absence of this validity and then learns, thereby improving the accuracy of subsequent designs. From the next design onwards, designs will be carried out in descending order of weighting, further improving the accuracy of the design. In other words, this optimization results in the design of more realistic and valid architectural drawing data 200, making it possible to design more realistic architectural drawings that satisfy the client's requests.

[0107] With this configuration, architectural drawing data 200 for a building that meets the client's requirements can be designed in one step, from the land to the room layout, making it possible to efficiently design architectural drawings that meet the client's requests with high precision.

[0108] (Third embodiment of the present invention) An architectural drawing design device 10 according to this embodiment will be described with reference to Fig. 8. As in the first embodiment, this device comprises the data management unit 11, the design unit 12, and the output control unit 13, and further, the data management unit receives and manages input of map data 101 relating to surveyed land, and the design unit includes a site determination unit 12e that identifies a site in the map data 101 managed by the data management unit as a building site based on at least the Building Standards Act and the City Planning Act.

[0109] This map data 101 is data based on map information from a public map that describes the surveyed shape of land, plots, lot numbers, roads, waterways, directions, etc. This map data 101 can be, for example, image data saved in PDF format.

[0110] The processing content added to the first embodiment will be described below. Fig. 8 is a diagram showing the processing of the design unit 12 of the architectural drawing design device 10 according to this embodiment.

[0111] As shown in FIG. 8, the data management unit 11 receives and manages input of map data 101 relating to surveyed land.

[0112] As shown in the memory flow of FIG. 8, the site determination unit 12e identifies the site shown in the map data 101 as the building site A, based on at least the Building Standards Act and the City Planning Act. While it has been difficult to uniquely identify the actual site for construction based on map information from a public map alone, the site determination unit 12e can identify the actual site for construction as the building site A by determining, for example, the distance from the road required, whether a slope is permitted on the site, and so on, based on at least the Building Standards Act and the City Planning Act. For this determination, criteria for the determination can be registered in advance in the map data 101, or these criteria can be learned and updated as default values ​​using the neural network of this embodiment. The design unit 12 then determines the building site A as the target site for the building, and the same process as in the first embodiment proceeds.

[0113] In this way, the data management unit 11 receives and manages the input of map data 101 relating to the surveyed land, and the design unit includes a land determination unit 12e that identifies the land in the map data 101 managed by the data management unit 11 as a building site based on at least the Building Standards Act and the City Planning Act. Therefore, the design is based on land where construction is actually possible, which increases the accuracy of the design, and enables the client's requests to be met while highly realistic architectural drawings to be designed.

[0114] (Fourth embodiment of the present invention) An architectural drawing design device 10 according to this embodiment will be described with reference to Fig. 9. As in the first embodiment, the device comprises the data management unit 11, the design unit 12, and the output control unit 13, and the design unit 12 further includes a zoning consistency determination unit 12f that determines the consistency of the zoning elements determined by the zoning element determination unit 12c as a custom-built house, and if the zoning consistency determination unit 12f determines that the zoning elements do not satisfy the consistency, the zoning element determination unit 12c is executed again to determine new zoning elements.

[0115] The processing content added to the first embodiment will be described below: Fig. 9 is a diagram showing the processing of the design unit 12 of the architectural drawing design device 10 according to this embodiment.

[0116] As shown in Figure 9, similar to the first embodiment, the zoning element judgment unit 12c judges the zoning elements including the sections related to the building based on the input client characteristic data 100 and the volume elements judged by the volume element judgment unit 12b, based on the zoning-related information, and based on the trained neural network.

[0117] The zoning consistency determining unit 12f can determine the consistency of the zoning elements as a building of a custom-built house based on, for example, custom-built house specific data 102 that specifies the characteristics of the custom-built house.

[0118] This custom-built home specific data 102 can be composed of custom-built home specific information that defines priorities for designing areas suitable for custom-built homes, such as larger bathrooms, spacious living / dining / kitchen areas, open-ceiling spaces, etc., without regard to the theories and constraints of prefabricated homes, such as bathroom size and living / dining / kitchen area, etc. This custom-built home specific information can use pre-registered criteria, and can also be updated by learning such criteria using the neural network of this embodiment.

[0119] If the zoning consistency determination unit 12f determines that the zoning elements are consistent with the building of a custom-built house, the same processing as in the first embodiment will proceed. If the consistency is not determined, the zoning element determination unit 12c will determine the zoning elements including the section related to the building until the zoning consistency determination unit 12f determines the consistency.

[0120] In this way, the zoning elements are repeatedly evaluated until a plot that is consistent with the requirements for a custom-built home is found. This allows the client's requests to be met by taking advantage of the unique characteristics of custom-built homes, which are different from pre-built homes, and allows architectural drawings to be designed that will increase customer satisfaction.

[0121] (Fifth embodiment of the present invention) An architectural drawing design device 10 according to this embodiment will be described with reference to Fig. 10. As in the first embodiment, the device includes the data management unit 11, the design unit 12, and the output control unit 13. The design unit 12 further includes a planning consistency determination unit 12g that determines the consistency of the planning elements determined by the planning element determination unit 12d as a custom-built house, and when the planning consistency determination unit 12g determines that the planning elements do not satisfy consistency, the zoning element determination unit 12c and the planning element determination unit 12d are executed again to determine new planning elements.

[0122] The processing content added to the first embodiment will be described below: Fig. 10 is a diagram showing the processing of the design unit 12 of the architectural drawing design device 10 according to this embodiment.

[0123] As shown in Figure 10, similar to the first embodiment, the planning element judgment unit 12d judges the planning elements based on the input client characteristic data 100 and the zoning elements judged by the zoning element judgment unit 12c, based on the planning-related information, and based on the trained neural network.

[0124] The planning consistency determining unit 12g can determine the consistency of the planning elements as a building of a custom-built house based on, for example, custom-built house specific data 102 that specifies the characteristics of the custom-built house.

[0125] This custom-built home-specific data 102 can be composed of custom-built home-specific information that defines priorities for designing rooms (floor plans) with a high degree of freedom and uniqueness suited to custom-built homes, such as separate bathrooms, island kitchens, sunken living rooms, home theaters, wood decks, raised ceilings, wall storage, telework spaces, built-in garages, living room staircases, and courtyards, while eliminating the theories and constraints of prefabricated homes, such as common use of bathrooms and kitchens and the use of standard kitchens. It can also define the degree of connection between adjacent rooms (floor plans), such as whether the kitchen is next to the living room or whether there are staircases near the entrance. This custom-built home-specific information can use pre-registered criteria, or it can be updated by learning such criteria as default values ​​using the neural network of this embodiment. When custom-built home-specific information is updated through learning, it becomes possible to incorporate and reflect the current trends in custom-built homes, allowing for the design of always-updated architectural drawings, further increasing customer satisfaction.

[0126] If the planning consistency determination unit 12g determines that the planning elements are consistent with the building of a custom-built house, the process proceeds in the same manner as in the first embodiment. If the consistency is not determined, the process returns to the process of the zoning element determination unit 12c, and the zoning element determination unit 12c and the planning element determination unit 12d determine the planning elements, including the floor plan, related to the building, until the planning consistency determination unit 12f determines the consistency.

[0127] In this way, the planning element judgment unit 12d repeatedly judges the planning elements until consistency as a custom-built home is achieved, making it possible to meet the client's requests with content that takes advantage of the unique characteristics of custom-built homes, which are different from prefabricated homes, and to design architectural drawings that will provide greater customer satisfaction.

[0128] As an example of the output of each determination unit of the architectural drawing design device according to each of the above embodiments, for example, as shown in Fig. 11(a), the external element determination unit 12a determines the target of external elements including the client's budget as area X. Next, for this area X, the volume element determination unit 12b determines the volume element including the site area of ​​the client's building as area Y, as shown in Fig. 11(b).

[0129] Next, as shown in Fig. 11(c), the zoning element determination unit 12c determines zoning elements including sections related to the client's building by dividing this area Y. Next, as shown in Fig. 11(d), the planning element determination unit 12d determines planning elements including floor plans related to the client's building for this divided area, and outputs them using, for example, numerical values ​​(numbers).

[0130] These numerical values ​​(numbers) are associated with respective meanings based on the table below, and can be converted into meaningful symbols (mnemonic codes) and displayed, such as "l" for living room and "d" for dining room, as shown in Figure 11(e). Based on this result, meaningful consecutive sections are grouped together, and the consecutive areas of "l" are treated as one living room, and output as a design drawing, as shown in Figure 11(f). Note that this design drawing can also be generated more graphically using generation AI, etc.

[0131] [Table 1]

[0132] (Sixth embodiment of the present invention) The architectural drawing design device 10 according to this embodiment will be described with reference to Fig. 12. As in the first embodiment, the device comprises the data management unit 11, the design unit 12, and the output control unit 13, and further, the client characteristic data 100 is formed as a distribution element made up of each of the architectural elements.

[0133] The distribution elements are obtained by displaying the client characteristic data 100 as, for example, a radar chart, as shown in Fig. 12. The radar chart allows the client characteristics to be visually grasped.

[0134] In this way, since the client characteristic data 100 is formed as a distribution element made up of each of these architectural elements, the client's characteristics can be grasped instantly and accurately using this distribution element, which increases the client's reliability in the architectural drawing data 200 according to their characteristics and further increases customer satisfaction.

[0135] As an application example, for example, as shown in FIG. 12, when there are multiple clients, the design unit 12 uses a trained neural network to design architectural drawing data 200 for a building that meets the requirements of the multiple clients, based on multiple distribution elements corresponding to multiple client characteristic data 100.

[0136] As shown in FIG. 12, for example, if the clients are a married couple, they may have different requirements and design different architectural drawing data 200 for each of them. Even in this case, the design unit 12 averages the radar charts of the client characteristic data 100, which are input data, and uses the averaged radar chart as input data to a trained neural network to generate a single, more optimal architectural drawing data 200. It is also possible to configure the output control to simultaneously output these two sets of architectural drawing data 200 for each client. This allows information to be provided for discussing the requirements of both parties, facilitating a smooth agreement between the parties on a custom-built home or other building.

[0137] In this way, even if there are multiple clients, it does not take a lot of time and effort to consolidate the many different requests of each client into one, and architectural drawing data 200 that consolidates the requests of each client can be efficiently designed in a shorter time, thereby further increasing customer satisfaction through time-efficient design.

[0138] The above-described embodiments may be configured independently or in combination with each other. For example, the third and fourth embodiments may be combined. The present invention will be described in more detail below with reference to examples, but the present invention is not limited to the following examples.

[0139] Example 1 Client characteristic data was obtained from the subject, Client A, showing the client's characteristics regarding architectural elements, as shown in Figure 13. This client characteristic data was obtained by inputting that Client A is a family of four with a junior child, that they need a Japanese-style room for the baby, that they want to park two cars, that their budget for the building is 20 million yen, and that they are cost-conscious.

[0140] Similar to the processing flow of Figure 11 above, the design unit of this architectural drawing design device designed a two-story building as shown in Figure 14, taking into consideration the fact that the client was a four-person junior family with children, the budget for the building was 20 million yen, and cost was a major consideration. Figure 14(a) is the floor plan for the first floor, which was generated using the zoning and numbers in Figure 14(b). Similarly, Figure 14(c) is the floor plan for the second floor, which was generated using the zoning and numbers in Figure 14(d).

[0141] Furthermore, based on these results, the design unit of this architectural drawing design device processed the first floor layout shown in Figure 15(a) and the second floor layout shown in Figure 15(b) in a more graphical manner. In the first floor layout shown in Figure 15(a), a parking space that can accommodate two cars was provided in front of the entrance. Also, as shown in Figure 15(a), a space for a baby was designed as a Japanese-style room adjacent to the living, dining, and kitchen area.

[0142] Example 2 From the subject client B, client characteristic data was obtained that shows the client's characteristics regarding architectural elements, as shown in Figure 16. This client characteristic data was obtained by inputting that client B is a family of four, that the husband's hobby is cars, that they would like to park two cars, that their budget for the building is 30 million yen, and that they are cost-conscious.

[0143] Similar to the processing flow of Figure 11 above, the design unit of this architectural drawing design device designed a two-story building as shown in Figure 17, taking into consideration a family of four, a budget of 30 million yen for the building, and cost as a priority. Figure 17(a) is the floor plan for the first floor, which was generated using the zoning and numbers in Figure 17(b). Similarly, Figure 17(c) is the floor plan for the second floor, which was generated using the zoning and numbers in Figure 17(d).

[0144] Furthermore, based on these results, the design unit of this architectural drawing design device processed the first floor floor plan shown in Figure 18(a), the second floor floor plan shown in Figure 18(b), and the loft floor plan shown in Figure 18(c) in a more graphical manner. For the first floor floor plan shown in Figure 18(a), a built-in garage was designed so that the cars could be viewed from the living, dining, and kitchen area, as the husband's hobby was cars and he wanted to park two cars. Furthermore, as shown in Figure 18(c), a loft was designed within the budget to accommodate the desire for ample storage space.

[0145] Example 3 From the subject, client C, we obtained client characteristic data that shows the client's characteristics regarding architectural elements, as shown in Figure 16. This client characteristic data was obtained by inputting that client C is a family of four with a junior child, that they want to wash their hands immediately after returning home, that they want to be able to park three cars in case they have guests, that they are cost-conscious with a budget of 20 million yen for the building, and that the husband is an avid outdoor enthusiast.

[0146] Similar to the processing flow of Figure 11 above, the design unit of this architectural drawing design device designed a two-story building as shown in Figure 20, taking into consideration the fact that the client was a four-person junior family with children, the budget for the building was 20 million yen, and cost was a major consideration. Figure 20(a) is the floor plan for the first floor, which was generated using the zoning and numbers in Figure 20(b). Similarly, Figure 20(c) is the floor plan for the second floor, which was generated using the zoning and numbers in Figure 20(d).

[0147] Furthermore, based on these results, the design unit of this architectural drawing design device processed the first floor floor plan shown in Figure 21(a) and the second floor floor plan shown in Figure 21(b) in a more graphical manner. For the first floor floor plan shown in Figure 21(a), a large garden with a wooden deck suitable for barbecues and other activities was designed within the budget, as the husband is an avid outdoor enthusiast. In addition, the washroom and dressing room were designed to be separate, as the owner wanted to be able to wash his hands immediately after returning home.

[0148] From the above examples, it was confirmed that custom-built homes that meet the requests of each of the clients A to C can be automatically designed. [Explanation of symbols]

[0149] 10. Architectural drawing design device 11 Data Management Department 11a Check processing section 12 Design Department 12A External element input content preprocessing 12a External element judgment section 12B External Element Optimization Section 12C Volume element input content preprocessing 12b Volume element determination unit 12D Volume Element Optimization Department 12E Zoning element input content preprocessing 12c Zoning Element Judgment Division 12F Zoning Element Optimization Department 12G Zoning Element Input Content Preprocessing 12d Planning Element Judgment Department 12H Planning Element Optimization Department 12e Site Judgment Department 12f Zoning Consistency Judgment Department 12g Planning Consistency Judgment Unit 13 Output control section 100 Client characteristic data 101 Map Data 102 Custom-built home specific data 200 architectural drawing data

Claims

1. An architectural drawing design program for designing architectural drawing data of a client, On the computer, a data management function for receiving and managing input of client characteristic data indicating client characteristics for a plurality of direct or indirect architectural elements; a design function for designing architectural drawing data for a building that meets the requirements of a client, based on the client characteristic data received from the data management function, using a trained neural network that has learned the interrelationships of each architectural element in the client characteristic data for past clients and architectural drawing data; an output control function for controlling the output of the architectural drawing data as data learned by a trained neural network; An architectural drawing design program characterized by realizing the above.

2. 2. The architectural drawing design program according to claim 1, The data management function receives and manages input of map data relating to the surveyed land, The design function includes a land use determination function that identifies a site in the map data managed by the data management function as a building site based on at least the Building Standards Act and the City Planning Act. An architectural drawing design program characterized by:

3. 2. The architectural drawing design program according to claim 1, The data management function determines the validity of the content of each architectural element of the client characteristic data that has been input. An architectural drawing design program characterized by:

4. 2. The architectural design program according to claim 1, The data management function determines the priority of each architectural element based on the interrelationship of each architectural element in the input client characteristic data. An architectural drawing design program characterized by:

5. 2. The architectural drawing design program according to claim 1, The design function optimizes the architectural drawing data generated by the trained neural network to satisfy content validity. An architectural drawing design program characterized by:

6. 2. The architectural drawing design program according to claim 1, The design function is an external element determination function for determining external elements including the client's budget; and a volume element determination function for determining volume elements including the site area of ​​the client's building based on the external elements; a zoning element determination function that determines a zoning element including a section related to the building of the client based on the volume element; a planning element determination function that determines planning elements including a floor plan related to the building of the client based on the zoning elements; Architectural drawing design program comprising:

7. 7. The architectural drawing design program according to claim 6, The design function is a planning consistency determination function that determines the consistency of the planning elements determined by the planning element determination function as a building of a custom-built house; If the planning consistency determination function determines that consistency is not satisfied, the zoning element determination function and the planning element determination function are executed again to determine new planning elements. An architectural drawing design program characterized by:

8. 7. The architectural drawing design program according to claim 6, The design function is A zoning consistency determination function is included that determines the consistency of the zoning elements determined by the zoning element determination function as a building for a custom-built house, If the zoning consistency determination function determines that consistency is not satisfied, the zoning element determination function is executed again to determine a new zoning element. An architectural drawing design program characterized by:

9. 2. The architectural design program according to claim 1, On the computer, A virtual space construction function for constructing a virtual space based on architectural drawing data designed by the design function is realized, The output control function controls the output of the virtual space. An architectural drawing design program characterized by:

10. 2. The architectural design program according to claim 1, The client characteristic data is formed as a distribution element consisting of each of the architectural elements. An architectural drawing design program characterized by:

11. The architectural design program according to claim 10, If there are multiple owners, The design function uses a trained neural network to design architectural drawing data for a building that meets the requirements of the plurality of clients, based on the plurality of distribution elements corresponding to the plurality of client characteristic data. An architectural drawing design program characterized by:

12. An architectural drawing design device that designs architectural drawing data of a client, a data management means for receiving and managing input of client characteristic data indicating client characteristics for a plurality of direct or indirect architectural elements; a design means for designing architectural drawing data for a building that meets the requirements of a client, using a trained neural network that has learned the interrelationships of each architectural element in the client characteristic data related to past clients and architectural drawing data based on the client characteristic data received from the data management means; output control means for controlling the output of the architectural drawing data as data learned by a trained neural network; An architectural drawing design device comprising:

13. 13. The architectural drawing design device according to claim 12, The data management means receives and manages input of map data relating to the surveyed land, The design means includes a site determination means for identifying a site in the map data managed by the data management means as a construction site based on at least the Building Standards Act and the City Planning Act. An architectural drawing design device characterized by:

14. 13. The architectural drawing design device according to claim 12, The data management means judges the validity of the content of each architectural element of the input client characteristic data. An architectural drawing design device characterized by:

15. The architectural design system according to claim 12, The data management means, based on the interrelationships of the architectural elements of the input client characteristic data, Determine the priority of each building element An architectural drawing design device characterized by:

16. 13. The architectural drawing design device according to claim 12, The design means optimizes the architectural drawing data generated by the trained neural network so as to satisfy the validity of the content. An architectural drawing design device characterized by:

17. 13. The architectural drawing design device according to claim 12, The design means an external factor determining means for determining external factors including the client's budget; a volume element determining means for determining a volume element including a site area related to the building of the client based on the external elements; a zoning element determination means for determining a zoning element including a section related to the building of the client based on the volume element; a planning element determining means for determining a planning element including a floor plan related to the building of the client based on the zoning element; An architectural drawing design device comprising:

18. 18. The architectural drawing design device according to claim 17, The design means a planning consistency determination means for determining the consistency of the planning elements determined by the planning element determination means as a building of a custom-built house; If the planning consistency determination means determines that consistency is not satisfied, the zoning element determination means and the planning element determination means are executed again to determine new planning elements. An architectural drawing design device characterized by:

19. 18. The architectural drawing design device according to claim 17, The design means A zoning consistency determination means is provided for determining the consistency of the zoning elements determined by the zoning element determination means as a building for a custom-built house, If the zoning consistency determination means determines that consistency is not satisfied, the zoning element determination means is executed again to determine new zoning elements. An architectural drawing design device characterized by:

20. The architectural design system according to claim 12, a virtual space construction means for constructing a virtual space based on the architectural drawing data designed by said design means, The output control means controls the output of the virtual space. An architectural drawing design device characterized by:

21. The architectural design system according to claim 12, The client characteristic data is formed as a distribution element consisting of each of the architectural elements. An architectural drawing design device characterized by:

22. The architectural design system according to claim 12, If there are multiple owners, The design means designs architectural drawing data for a building that meets the requirements of the plurality of clients using a trained neural network based on the plurality of distribution elements corresponding to the plurality of client characteristic data. An architectural drawing design device characterized by:

Images