Building design scheme automatic generation method and system based on artificial intelligence

Abstract

The invention discloses an automatic building design scheme generation method and system based on artificial intelligence, and relates to the technical field of intelligent building management, the system comprises a control center, and the control center is connected with a front end acquisition module, a structure processing module, a function analysis module and a construction design module; collecting design preparation data and user demand information, and constructing a building associated space; performing category division on the design preparation data to obtain design feature categories, constructing a step constraint information set according to the design feature categories, and performing keyword screening on the user demand information to obtain user demand keywords; performing similarity screening on the step constraint information set according to user demand keywords and constructing a demand selectable database; performing scheme construction on the demand selectable database to obtain a primary design scheme, and performing user simulation pushing on the primary design scheme through the building associated space to obtain an optimal pushing scheme; the design efficiency is improved, the design quality is optimized, and personalized design is realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of intelligent building management, and particularly relates to a building design scheme automatic generation method and system based on artificial intelligence. BACKGROUND

[0002] In the traditional building design process, designers need to comprehensively consider building specifications, functional requirements, aesthetic principles and other aspects to design a building scheme that meets practical requirements and has artistic beauty. However, this process often takes a long time and requires designers to have rich experience and deep professional knowledge. In addition, with the increasing demand for individualization and customization in the construction industry, designers face greater challenges.

[0003] With the rapid development of technology, artificial intelligence technology is increasingly widely used in various fields, and the construction industry has ushered in a digital and intelligent transformation. In the field of building design, the introduction of artificial intelligence technology provides designers with more efficient and innovative design schemes. In order to further improve design efficiency, reduce design cost, and at the same time realize more innovative and diversified design schemes, the building design scheme automatic generation method based on artificial intelligence emerges as the times require. This method uses artificial intelligence technology to learn and analyze a large amount of building data to automatically generate building design schemes that meet specific requirements, not only improving the efficiency and quality of design work, but also promoting the digital transformation of the construction industry and bringing more possibilities to architects and users. SUMMARY

[0004] The building design scheme automatic generation system based on artificial intelligence comprises a control center, wherein the control center is connected with a front-end acquisition module, a structure processing module, a function analysis module and a construction design module. The front-end acquisition module is used for acquiring design preparation data and user demand information, and constructing a building associated space. The structure processing module is used for classifying the design preparation data, obtaining design feature categories, setting quantitative scoring standards to score the design feature categories, obtaining user feature category scores, constructing step constraint information set according to the design feature categories, and screening key words from the user demand information, and obtaining user demand key words. The function analysis module is used for screening the step constraint information set according to the user demand key words, obtaining initial matching similarity, pre-constructing design information from the initial matching similarity, and obtaining a demand optional database. The construction design module is used for constructing schemes from the demand optional database, obtaining a preliminary design scheme, scoring the preliminary design scheme based on the user feature category scores, obtaining a preliminary scheme evaluation score, simulating and pushing the preliminary design scheme through the building associated space, and obtaining an optimal pushing scheme.

[0005] Preferably, the process of collecting design preparation data and user requirement information includes: Information is collected for construction projects to obtain design preparation data; Set up a user data collection terminal to collect user needs information from target users. Based on user demand information, regional data collection for building construction is conducted to obtain target construction blocks, and building-related spaces are constructed based on these target construction blocks.

[0006] Preferably, the process of obtaining user feature category scores includes: The design preparation data is classified to obtain design feature categories, and the information of the design feature categories is summarized to obtain feature category information; Quantitative scoring criteria are set based on the obtained design feature categories; The obtained design feature categories and quantitative scoring standards are uploaded to the Architecture Companion Space. The Architecture Companion Space evaluates the target users according to the obtained quantitative scoring standards and obtains user feature category scores.

[0007] Preferably, the process of obtaining user demand keywords includes: Obtain design feature categories, perform constraint statistics on design feature categories, obtain a step constraint information set, and upload the obtained design feature categories to the step constraint information set; Based on the step-constraint information set, feature analogy extraction is performed on user requirement information according to the design feature category to obtain user requirement keywords.

[0008] Preferably, the process of similarly filtering the step constraint information set based on user demand keywords includes: Set retrieval ports for the step constraint information set, and set the order according to the design feature categories to obtain the retrieval feature order; Based on the search feature order, user demand keywords are uploaded to the search port. The search port then filters target users' needs based on these keywords to obtain initial matching similarity.

[0009] Preferably, the process of preconstructing design information for the initial matching similarity includes: A threshold is selected for the initial matching similarity to obtain the initial matching feature segment sequence; The obtained initial matching feature segment sequence is matched with the design feature category, and the successfully matched initial matching feature segment sequence is associated with the design feature category. A raw requirement database is constructed based on design feature categories. The obtained initial matching feature segment sequence is uploaded to the raw requirement database to obtain a requirement optional database.

[0010] Preferably, the process of obtaining the preliminary design scheme includes: Based on the design feature categories, a framework is selected from the optional database of requirements to obtain the initial framework of the solution; Based on the initial framework of the solution, the database of optional requirements is supplemented to obtain a preliminary design solution; Obtain user feature category scores, and based on the user feature category scores, calculate the scores of the preliminary design schemes according to the design feature categories to obtain the preliminary scheme evaluation scores. Then, associate the obtained preliminary scheme evaluation scores with the corresponding preliminary design schemes.

[0011] Preferably, the process of simulating and pushing the initial design scheme to users through the building's associated space includes: The preliminary design schemes are uploaded to the architectural companion space, and the virtual effect of the preliminary design schemes is displayed through the architectural companion space to obtain the scheme effect model. The obtained preliminary design scheme evaluation score is marked on the corresponding scheme effect model. By using the building's associated space to push the effect model of the proposed scheme to users, the target users can make personalized judgments based on the effect model of the pushed scheme according to the initial scheme evaluation score, and obtain the best pushed scheme.

[0012] Based on the aforementioned AI-based automatic architectural design scheme generation system, this invention also provides an AI-based automatic architectural design scheme generation method, comprising the following steps: Step 1: Collect design preparation data and user needs information to construct the building-related space; Step 2: Categorize the design preparation data to obtain design feature categories, set quantitative scoring criteria to conduct user scoring of design feature categories, obtain user feature category scores, construct a set of step constraint information based on design feature categories, and filter user demand information by keywords to obtain user demand keywords; Step 3: Based on the user's requirement keywords, perform similarity filtering on the step constraint information set to obtain the initial matching similarity. Then, pre-construct design information based on the initial matching similarity to obtain the requirement optional database. Step 4: Construct a solution from the database of available requirements to obtain preliminary design solutions. Score the preliminary design solutions based on user characteristic category ratings to obtain preliminary solution evaluation scores. Simulate user push of the preliminary design solutions through the building's associated space to obtain the optimal push solution.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. Collect historical construction project data within the area to be constructed, extract project construction feature categories, and obtain personalized scores from target users based on the feature information corresponding to the construction feature categories. This will help to obtain users' preferences for different construction features, making it easier to develop personalized design schemes and meet users' personalized needs for architectural space. 2. By combining the extracted construction feature categories with the collected user demand information, the user demand keywords can be obtained, which can more accurately understand the user's specific needs and generate architectural solutions that better meet the user's requirements. 3. Simultaneously, a dataset is constructed from the obtained construction feature categories to obtain a set of step constraint information. Key information is matched against the set of step constraint information using user requirement keywords. This enables the automatic filtering and combination of design elements, reducing the complexity and time cost of manual design. Furthermore, the matched information is combined to obtain an initial matching scheme, ensuring that the generated scheme meets the user's basic requirements and improving the matching degree and practicality of the scheme. 4. The initial matching scheme is virtually displayed using the constructed virtual building space. Users can see the design scheme intuitively before actual construction, providing a more intuitive and interactive experience. It also allows users and designers to discover potential problems in time, reducing errors and rework during actual construction. Finally, a comprehensive score is obtained by combining the user's rating of the construction feature categories to select the best building scheme. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the present invention. Detailed Implementation

[0016] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] like Figure 1 As shown, the AI-based automatic architectural design scheme generation system includes a control center, which is connected to a front-end acquisition module, a structural processing module, a functional analysis module, and a construction design module. The front-end acquisition module is used to collect design preparation data and user demand information to construct building-related spaces. The structure processing module is used to classify the design preparation data to obtain design feature categories, set quantitative scoring standards to score the design feature categories by users, obtain user feature category scores, construct a set of step constraint information based on the design feature categories, and filter user demand information by keywords to obtain user demand keywords. The functional analysis module is used to perform similarity filtering on the step constraint information set based on user demand keywords, obtain initial matching similarity, pre-construct design information on the initial matching similarity, and obtain a demand optional database. The construction design module is used to construct schemes from the database of available requirements, obtain preliminary design schemes, score the preliminary design schemes based on user characteristic category scores, obtain preliminary scheme evaluation scores, and simulate user push of the preliminary design schemes through the building's associated space to obtain the optimal push scheme.

[0018] In practical applications, traditional architectural design processes require significant time and human resources, consuming substantial amounts of capital and struggling to generate innovative solutions. Artificial intelligence (AI) technology can enhance innovation in the design process, meet personalized user needs, and generate customized solutions. Furthermore, intelligent design processes can reduce human error and improve design performance through algorithmic optimization. This approach maintains design quality and creativity while increasing efficiency and responsiveness. The initial step involves a front-end data acquisition module to gather the necessary design preparation data for architectural design. The specific process includes: Information is collected for construction projects to obtain design preparation data; It should be further explained that, in the specific implementation process, the construction project refers to all construction projects within the construction area that need to be carried out. These are projects that have been established before the target design scheme is generated and can provide data support, including but not limited to the construction requirements, design specifications, and construction policies of the area. The information collection refers to the collection of preparatory data needed to generate the required architectural design scheme, i.e., design preparation data, including but not limited to historical project data, design specification requirements, user needs, geographical environment information, market data, and technical data. Among them, historical project data refers to past architectural project cases, such as design drawings, construction drawings, and project reports; design specification requirements refer to architectural design codes, safety standards, etc.; user needs refer to the needs of project owners and target users, such as spatial functions, comfort requirements, and budget constraints; geographical environment information refers to the climate, topography, and environmental characteristics of the project location; market data refers to building material prices, labor costs, and market trends; and technical data refers to new building materials, building technologies, and construction methods.

[0019] Set up a user acquisition terminal to collect the needs of target users, obtain user demand information, and associate the obtained user demand information with the corresponding target users. The target user refers to a client who needs architectural design services. They are clients who provide requirements and generate architectural design schemes based on those requirements. The collected requirement data is the user requirement information of the target user, including but not limited to the construction area, function, style, scale, cost, and sustainability requirements of the building.

[0020] Based on user demand information, regional data collection for building construction is performed to obtain the target construction area; The area collection refers to extracting building construction area information based on user demand information, which is the target construction block, representing the building area where the target user needs to carry out construction.

[0021] Based on the target construction block, a building-related space is constructed. The building-related space represents a virtual blank space used to display the simulated design scheme of the target construction block, which is exactly the same as the location and function of the target construction block in reality.

[0022] The obtained target construction block is uploaded to the building companion space, and the target construction block is transformed into a twin construction block through the building companion space. The twin transformation refers to creating a twin of the uploaded target construction block in the building's associated space, resulting in a twin virtual block with the same structure and function as the real-world target construction block, which is the twin construction block.

[0023] The structure processing module is used to classify design preparation data to obtain design feature categories, set quantitative scoring standards to evaluate user scores for each design feature category, construct a set of step constraint information based on the design feature categories, and filter user demand information using keywords to obtain user demand keywords. The specific process includes: The obtained design preparation data is classified to obtain design feature categories; It should be further explained that, in the specific implementation process, the data classification refers to the categorization of design preparation data to obtain different types in the design scheme, which are the design feature categories. For example, based on design preparation data including but not limited to historical project data, design specification requirements, user needs, geographical environment information, market data, and technical data, the categorized design feature categories include, but are not limited to, building type, building floors, building area, building structure type, design style, appearance features, interior layout, building materials used, construction technology, construction cost, construction budget, construction geographical location, and climate conditions. Each design feature category has corresponding preparation data information. For example, the design feature category of building type includes residential buildings, commercial buildings, industrial buildings, transportation buildings, and entertainment buildings, and the design feature category of design style includes modern style, traditional style, minimalist style, rural style, industrial style, and ecological style.

[0024] The obtained design feature categories are summarized to obtain feature category information, and the obtained feature category information is associated with the corresponding design feature categories; The information summarization refers to matching the design information corresponding to each category according to the divided design feature categories, which is the feature category information; for example, for the design feature category of building type, the corresponding feature category information includes residential buildings, commercial buildings, industrial buildings, transportation buildings, and entertainment buildings. Accordingly, these feature category information include basic information such as specific building models and construction rules for residential buildings, commercial buildings, industrial buildings, transportation buildings, and entertainment buildings.

[0025] Quantitative scoring criteria are set based on the obtained design feature categories; The quantitative scoring standard refers to the scoring criteria set for target users to judge the satisfaction of each design feature category. Satisfaction represents the target user's degree of liking or satisfaction with the design feature category. For example, different degrees of liking are given to modern style, traditional style, minimalist style, country style, industrial style, and eco style, and then there are corresponding score values. The set quantitative scoring standard is a consistent and standardized scoring criterion. In this embodiment, a scoring standard with a maximum score of 10 is used for evaluation. For example, the target user's score values ​​for modern style, traditional style, minimalist style, country style, industrial style, and eco style are 8, 2, 6, 1, 2, and 7, respectively. In particular, there is a corresponding quantitative scoring standard for each design feature category.

[0026] The obtained design feature categories and quantitative scoring standards are uploaded to the building companion space. The building companion space evaluates the target users according to the obtained quantitative scoring standards, obtains user feature category scores, and marks the obtained user feature category scores in the corresponding feature category information. It should be further explained that, in the specific implementation process, the category evaluation means that in the building-related space, the feature category information corresponding to each category of the design feature category is scored according to the set quantitative scoring standard, and the user feature category score corresponding to the feature category information is obtained. Each user feature category score is marked at the corresponding feature category information. That is, the target user's satisfaction score can be directly obtained through the feature category information, which is used to determine the target user's liking for different categories, so as to generate personalized architectural design schemes. After the category evaluation, each design feature has a target user's satisfaction score.

[0027] Obtain design feature categories, perform constraint statistics on the obtained design feature categories, obtain a step constraint information set, and upload the obtained design feature categories to the step constraint information set; The constraint statistics represent the construction of an information set, which uploads all design feature categories and their corresponding associated feature category information to this information set to obtain the step constraint information set. The step constraint information set includes all feature category information and the user feature category score after the target user performs category evaluation.

[0028] Based on the step constraint information set, feature analogy extraction is performed on user requirement information according to design feature categories to obtain user requirement keywords; The feature category extraction refers to extracting feature keywords from the content included in the user requirement information based on the feature categories of different design projects corresponding to the design feature categories in the constraint information set in the step, thereby obtaining user requirement keywords. For example, based on user requirement information including but not limited to construction area, building function, style, scale, cost, and sustainability requirements, the extracted user requirement keywords include construction scope restrictions, building function requirements, building style, building cost, and sustainability. This represents the target information needed to generate architectural design schemes for the target user. This allows for the acquisition of the target user's requirement characteristics, facilitating the generation of personalized design schemes. Furthermore, using design feature categories divided by design preparation data to help extract feature categories from user requirement information can improve the feature category matching speed, perform targeted category matching, and prevent the extraction of design feature categories not divided in the design preparation data from user requirement information. The design preparation data is collected from historical data of all construction projects in the construction area, which are project schemes that can be constructed, ensuring to the greatest extent possible that the design scheme generated for the target user is also feasible.

[0029] The functional analysis module is used to perform similarity filtering on the step constraint information set based on user requirement keywords, obtain initial matching similarity, pre-construct design information from the initial matching similarity, and obtain a database of optional requirements. The specific process includes: A retrieval port is set for the obtained set of step constraint information. The retrieval port is a port set in the set of step constraint information, which is used to place user feature information for matching. The matching object is the feature category information corresponding to the design feature category in the set of step constraint information. The obtained design feature categories are ordered to obtain the search feature order. The order setting refers to setting the search keyword order according to the order of the design feature categories, denoted as the search feature order. The search feature order is obtained by obtaining the user requirement keywords corresponding to the different categories in the design feature categories, thus obtaining the keyword search matching order of the target user. In particular, since the user requirement keywords are feature keywords extracted according to the design feature categories, determining the search feature order can determine the search order of the user requirement keywords. Therefore, even if the target user does not provide the corresponding design feature category keywords, the search feature order can be used to supplement the requirements, so that the search matches the design requirements that the target user truly needs. Based on the search feature order, the obtained user demand keywords are uploaded to the search port. The search port then filters the target users' needs based on the user demand keywords to obtain the initial matching similarity. It should be further explained that, in the specific implementation process, the requirement screening means that, at the retrieval port, the user requirement keywords of the target user are uploaded to the retrieval port in sequence according to the retrieval feature order, and only one user requirement keyword is uploaded at a time. The retrieval port matches the uploaded user requirement keywords with the design feature categories in the step constraint information set until all user requirement keywords corresponding to the retrieval feature order have been screened. In particular, if there is no corresponding user requirement keyword for a retrieval feature order, the user collection terminal collects supplementary features of the target user to obtain the missing user requirement keywords for the design feature categories of the retrieval feature order. This ensures that the final generated design scheme is generated based on all user requirements, rather than a scheme generated based on only a portion of the requirements that does not meet the target user's preferences. In this embodiment, natural language processing technology is used to match feature category information corresponding to design feature categories that are highly similar to user demand keywords. Since there is corresponding feature category information for each design feature category, the user demand keywords are matched with the design feature categories. That is, the user demand keywords are matched with the feature category information. The similarity of all feature category information that matches the user demand keywords in the step constraint information set is obtained and recorded as the initial matching similarity. Based on the different similarities of different feature category information, they can be sorted to obtain the feature category information with the highest similarity for scheme generation.

[0030] A threshold is selected for the obtained initial matching similarity to obtain the initial matching feature segment sequence; Furthermore, the threshold selection means setting a threshold for all initial matching similarities obtained from demand screening to obtain a similarity evaluation threshold. The similarity evaluation threshold represents the similarity threshold set for each user demand keyword. The portion of the initial similarities that is greater than the similarity evaluation threshold is selected and recorded as the initial matching segment, representing the feature category information that meets the matching degree. The feature category information that meets the initial matching similarity is sorted from largest to smallest according to the size of the similarity to obtain the initial matching feature segment sequence.

[0031] Based on the obtained design feature categories, the obtained initial matching feature segment sequences are categorized and matched. The successfully matched initial matching feature segment sequences are associated with the design feature categories. Here, category matching means that each initial matching feature segment sequence is obtained by selecting a threshold based on the fact that each user demand keyword has a corresponding design feature category. Therefore, each design feature category has a corresponding initial matching feature segment sequence after being matched with the target user's needs.

[0032] A requirement database is constructed based on design feature categories. The obtained initial matching feature segment sequences are uploaded to the requirement database to obtain a requirement optional database. The order of the initial matching feature segment sequences in the requirement database is the same as the order of the retrieved features.

[0033] The construction design module is used to construct schemes from the demand option database to obtain preliminary design schemes, score the preliminary design schemes based on user characteristic category scoring to obtain preliminary scheme evaluation scores, and simulate user push of the preliminary design schemes through the building's associated space to obtain the optimal push scheme. The specific process includes: Based on the obtained design feature categories, a framework is selected from the available requirement database to obtain the initial framework of the solution; The framework selection refers to selecting a feature category from the initial matching feature segment sequence corresponding to each design feature category in the demand optional database according to the order of feature retrieval, based on the obtained design feature categories, and forming the original framework structure of the scheme, which is the initial framework of the scheme. Here, "selecting a feature category" means randomly selecting a feature category from the initial matching feature segment sequence. In this embodiment, in order to improve the selection speed and avoid omissions, the selection is performed according to the order of the initial matching feature segment sequence. That is, the initial framework of the scheme is obtained by combining the first feature category information of all initial matching feature segment sequences.

[0034] Based on the initial framework of the solution, the obtained database of optional requirements is supplemented to obtain a preliminary design solution; It needs further clarification that, in the specific implementation process, the supplementary scheme indicates that, according to the framework structure of the initial scheme framework, the feature category information corresponding to the first design feature category in the initial scheme framework is selected and recorded as the variable feature item. This means that the feature category information in the initial scheme framework can be arbitrarily assigned as the variable feature item. Each time a scheme is generated, only one variable is changed, that is, only one design feature category is changed. Based on the feature category information in the initial matching feature segment sequence, each supplementary scheme only modifies the variable feature item to the next feature category information in the initial matching feature segment sequence, thus obtaining the preliminary design scheme. That is, the preliminary design scheme, compared to the initial scheme framework, only changes one feature category information. The process continues until the final feature category information of the initial matching feature segment sequence is reached. The feature category information corresponding to the first design feature category is then recorded as the changed feature item, and the feature category information corresponding to the second design feature category is recorded as the changed feature item. This process of obtaining the initial design scheme is repeated until all feature category information in the initial matching feature segment sequence is added to the initial scheme framework, and duplicate design schemes are deleted. Finally, all the initial design schemes are obtained. These are design schemes composed of feature category information of architectural projects with high similarity to user demand keywords. Further optimization and screening are needed to obtain the best design scheme to push to the target user.

[0035] Obtain user feature category scores, and based on the user feature category scores, calculate the scores of the preliminary design schemes according to the design feature categories to obtain the preliminary scheme evaluation scores. Then, associate the obtained preliminary scheme evaluation scores with the corresponding preliminary design schemes. The score statistics represent that for each design feature category included in each preliminary design scheme, there is a corresponding user feature category score. The user feature category scores of all feature category information are summed to obtain the comprehensive score of the preliminary design scheme, which is the preliminary scheme score. It represents the degree of liking of the newly generated design scheme obtained based on the target user's personalized scores of all feature categories of the existing architectural design project scheme, and is used to determine the degree of interest and preference of the newly generated design scheme. In particular, if the target user has special needs, that is, they value a certain feature category in the design solution more, then when obtaining the preliminary solution evaluation score, it is not only to sum the user feature category scores of all feature category information, but also to set weight ratios according to the importance of different feature categories, and calculate the corresponding total score based on the weight ratio, which is the preliminary solution evaluation score. This is conducive to meeting the reasonable needs and key concerns of the target user.

[0036] The obtained preliminary design schemes are uploaded to the architectural companion space. The architectural companion space displays the virtual effect of the preliminary design schemes to obtain the scheme effect model. The obtained preliminary design scheme evaluation score is marked on the corresponding scheme effect model. The virtual effect display refers to the process of transforming the obtained preliminary design scheme into a three-dimensional architectural design model in the building's associated space using virtual space. In this embodiment, three-dimensional modeling software is used to generate a three-dimensional rendering model of the building based on the preliminary design scheme in the building's associated space. This rendering model can directly display the architectural renderings of the design content and requirements corresponding to the preliminary design scheme.

[0037] By using the building's associated space to push the scheme effect model to users, the target users make personalized judgments based on the effect model of the pushed scheme according to the initial scheme evaluation score, and obtain the best pushed scheme. It should be further explained that, in the specific implementation process, the user push means that the generated scheme effect models are displayed to the target user in sequence in the building-related space. The target user selects the most favorite scheme effect model according to the displayed scheme effect models. Then, the initial design scheme corresponding to the scheme effect model is the design scheme most desired by the target user and is recorded as the best push scheme. Mark the corresponding preliminary design scheme and the preliminary scheme evaluation score at each scheme effect model, and sort all the preliminary scheme evaluation scores in descending order to obtain the evaluation score sequence; The model is displayed to the target user based on the obtained evaluation score sequence through the building-related space. The target user makes a personalized selection based on the displayed scheme effect model and obtains the best push scheme. The scheme effect model is displayed according to the initial selection scheme evaluation score. The initial design scheme corresponding to the scheme effect model that the target user selects is recorded as the best push scheme. In this embodiment, the initial design scheme corresponding to the first pushed scheme effect model is selected as the best push scheme. Specifically, if the target user does not select the initial design scheme corresponding to the first pushed solution effect model as the best push solution, they can choose their preferred design scheme as the best push solution based on their own preferences and the effect display of the solution effect model. In this case, the feature category information of the design feature category within the best push solution can be adjusted for optimization and upgrading. Only one feature category information is changed at a time until all feature category information is adjusted, resulting in a new set of adjusted design schemes. Combined with the solution effect model of the adjusted design scheme, a second push selection is performed. From the adjusted schemes, the preferred design scheme is obtained. During the second optimization and upgrading, the user's supplementary needs can be collected through the user collection terminal as the feature category information for adjustment, so as to further align with the design needs of the target user and obtain the final optimized preferred design scheme.

[0038] Based on the aforementioned AI-based automatic architectural design scheme generation system, this invention also provides an AI-based automatic architectural design scheme generation method, comprising the following steps: Step 1: Collect design preparation data and user needs information to construct the building-related space; Step 2: Categorize the design preparation data to obtain design feature categories, set quantitative scoring criteria to conduct user scoring of design feature categories, obtain user feature category scores, construct a set of step constraint information based on design feature categories, and filter user demand information by keywords to obtain user demand keywords; Step 3: Based on the user's requirement keywords, perform similarity filtering on the step constraint information set to obtain the initial matching similarity. Then, pre-construct design information based on the initial matching similarity to obtain the requirement optional database. Step 4: Construct a solution from the database of available requirements to obtain preliminary design solutions. Score the preliminary design solutions based on user characteristic category ratings to obtain preliminary solution evaluation scores. Simulate user push of the preliminary design solutions through the building's associated space to obtain the optimal push solution.

[0039] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An AI-based automatic architectural design scheme generation system, including a control center, characterized in that: The control center is connected to a front-end acquisition module, a structural processing module, a functional analysis module, and a construction design module. The front-end acquisition module is used to collect design preparation data and user demand information to construct building-related spaces. The structure processing module is used to classify the design preparation data to obtain design feature categories, set quantitative scoring standards to score the design feature categories by users, obtain user feature category scores, construct a set of step constraint information based on the design feature categories, and filter user demand information by keywords to obtain user demand keywords. The functional analysis module is used to perform similarity filtering on the step constraint information set based on user demand keywords, obtain initial matching similarity, pre-construct design information on the initial matching similarity, and obtain a demand optional database. The construction design module is used to construct schemes from the database of available requirements, obtain preliminary design schemes, score the preliminary design schemes based on user characteristic category scores, obtain preliminary scheme evaluation scores, and simulate user push of the preliminary design schemes through the building's associated space to obtain the optimal push scheme.

2. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of collecting design preparation data and user needs information includes: Information is collected for construction projects to obtain design preparation data; Set up a user data collection terminal to collect user needs information from target users. Based on user demand information, regional data collection for building construction is conducted to obtain target construction blocks, and building-related spaces are constructed based on these target construction blocks.

3. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of obtaining user characteristic category scores includes: The design preparation data is classified to obtain design feature categories, and the information of the design feature categories is summarized to obtain feature category information; Quantitative scoring criteria are set based on the obtained design feature categories; The obtained design feature categories and quantitative scoring standards are uploaded to the Architecture Companion Space. The Architecture Companion Space evaluates the target users according to the obtained quantitative scoring standards and obtains user feature category scores.

4. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of obtaining user demand keywords includes: Obtain design feature categories, perform constraint statistics on design feature categories, obtain a step constraint information set, and upload the obtained design feature categories to the step constraint information set; Based on the step-constraint information set, feature analogy extraction is performed on user requirement information according to the design feature category to obtain user requirement keywords.

5. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of filtering the step constraint information set based on user demand keywords includes: Set retrieval ports for the step constraint information set, and set the order according to the design feature categories to obtain the retrieval feature order; Based on the search feature order, user demand keywords are uploaded to the search port. The search port then filters target users' needs based on these keywords to obtain initial matching similarity.

6. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of preconstructing design information for initial matching similarity includes: A threshold is selected for the initial matching similarity to obtain the initial matching feature segment sequence; The obtained initial matching feature segment sequence is matched with the design feature category, and the successfully matched initial matching feature segment sequence is associated with the design feature category. A raw requirement database is constructed based on the design feature categories. The obtained initial matching feature segment sequence is uploaded to the raw requirement database to obtain a database of optional requirements.

7. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of obtaining the preliminary design scheme includes: Based on the design feature categories, a framework is selected from the optional database of requirements to obtain the initial framework of the solution; Based on the initial framework of the solution, the database of optional requirements is supplemented to obtain a preliminary design solution; Obtain user feature category scores, and based on the user feature category scores, calculate the scores of the preliminary design schemes according to the design feature categories to obtain the preliminary scheme evaluation scores. Then, associate the obtained preliminary scheme evaluation scores with the corresponding preliminary design schemes.

8. The automatic generation system for architectural design schemes based on artificial intelligence according to claim 1, characterized in that, The process of simulating user feedback on preliminary design schemes through the building's associated spaces includes: The preliminary design schemes are uploaded to the architectural companion space, and the virtual effect of the preliminary design schemes is displayed through the architectural companion space to obtain the scheme effect model. The obtained preliminary design scheme evaluation score is marked on the corresponding scheme effect model. By using the building's associated space to push the effect model of the proposed scheme to users, the target users can make personalized judgments based on the effect model of the pushed scheme according to the initial scheme evaluation score, and obtain the best pushed scheme.

9. The method for automatically generating architectural design schemes according to any one of claims 1 to 8, characterized in that, Includes the following steps: Step 1: Collect design preparation data and user needs information to construct the building-related space; Step 2: Categorize the design preparation data to obtain design feature categories, set quantitative scoring criteria to conduct user scoring of design feature categories, obtain user feature category scores, construct a set of step constraint information based on design feature categories, and filter user demand information by keywords to obtain user demand keywords; Step 3: Based on the user's requirement keywords, perform similarity filtering on the step constraint information set to obtain the initial matching similarity. Then, pre-construct design information based on the initial matching similarity to obtain the requirement optional database. Step 4: Construct a solution from the database of available requirements to obtain preliminary design solutions. Score the preliminary design solutions based on user characteristic category ratings to obtain preliminary solution evaluation scores. Simulate user push of the preliminary design solutions through the building's associated space to obtain the optimal push solution.

Images