A building design modeling system

Abstract

The present invention provides a building design modeling system, including core components such as a surveying and mapping module, an image description module, an image generation module, a model generation module, and a processing control module. The surveying and mapping data of the design plot is obtained by the surveying and mapping module, and the top view image of the plot and the surrounding scene image are extracted. Subsequently, the image description module is used to generate description texts of the plot features and the surrounding scene respectively. The processing control module combines these description texts with preset architectural design requirements to generate architectural design concept texts. Based on this concept text, the image generation module creates multiple architectural pre-construction drawings to form a building pre-construction image data set. Finally, the model generation module converts these pre-construction drawings into corresponding architectural pre-construction models. The automation of the design process is realized, and the plot features and surrounding environmental factors can be fully considered to generate multiple architectural design schemes that meet the actual scene, thereby improving the design efficiency and providing flexibility and selection space for architectural design.

Description

Technical Field

[0001] The present application relates to the field of architectural design, and in particular to an architectural design modeling system. Background Art

[0002] Architectural design modeling system is a computer-aided design system used for architectural design, planning and construction. Such systems can help architects and engineers create digital 3D models of buildings and realize digital management of the entire process from conceptual design to construction drawings.

[0003] Existing architectural design modeling systems have some significant limitations and deficiencies. The first problem is that the learning threshold of the software is high, and new users need to invest a lot of time to master its complex functions and operation methods. Secondly, in the creative design stage, these systems often limit the designer's freedom due to their standardized modeling methods, which is not conducive to the expression of conceptual creativity.

[0004] In the field of architectural design, these systems are still insufficient in dealing with non-standard components and complex geometric shapes, and it is difficult to meet the pursuit of free forms in contemporary architectural design. At the same time, the rendering effects and material expressiveness provided by the system are limited, and it is difficult to fully meet the artistic needs of architectural expression. Summary of the invention

[0005] The present application provides a building design modeling system, comprising: Surveying and mapping module; Image description module; Image generation module; Model generation module; Processing control module; Wherein, the surveying and mapping module, the image description module, the image generation module, and the model generation module are respectively data-connected to the processing control module; Wherein, the architectural design modeling system further includes an architectural design modeling strategy, including the following steps: A1, obtaining corresponding design plot area surveying data through surveying by the surveying module according to the preset building design plot and the preset surveying radius; A2, extracting corresponding land parcel top view image data and land parcel surrounding scene image data from the designed land parcel area surveying and mapping data through the processing control module; A3, generating corresponding land parcel feature description text through the image description module according to the land parcel overhead image data; A4, generating corresponding surrounding scene feature description text through the image description module according to the surrounding scene image data of the land parcel; A5, generating a corresponding architectural design concept text by combining the plot feature description text, the surrounding scene feature description text and the preset architectural design requirement text through the processing control module; A6, generating corresponding pre-constructed image data of each building through the image generation module according to the architectural design concept text and the preset number of image generation, and combining the pre-constructed image data of each building to generate a pre-constructed image data set of the building; A7, generating corresponding prefabricated building model data through the model generation module for each prefabricated building image data in the prefabricated building image data set.

[0006] By adopting the above technical solution, the architectural design modeling system can combine surveying and mapping data, image analysis, text description and model generation, and can simultaneously consider the characteristics of the plot and the surrounding environmental factors, so as to generate an architectural design solution that better meets the actual scene requirements. By generating a prefabricated architectural image data set, multiple design solutions are provided for selection, which increases the flexibility and selectivity of the design. The entire process is highly automated, which can significantly improve the efficiency of architectural design while ensuring the coordination of the design solution with the surrounding environment.

[0007] Optionally, the architectural design modeling system further comprises the following steps: A8, determining corresponding building model bottom surface image data according to each building prefabricated model data; A9, calculating the graphic matching degree of the corresponding building plot according to the bottom surface image data of the building model and the top view image data of the plot using a preset graphic matching algorithm; A10: If the matching degree of the building plot graphic is not less than a preset matching threshold, the corresponding prefabricated building model data is defined as available building design model data.

[0008] By adopting the above technical solution, the architectural design modeling system can perform a matching degree analysis between the building bottom surface image of the generated building prefabricated model data and the actual plot image to increase the model verification and screening mechanism, thereby being able to evaluate the adaptability of the architectural design scheme and the plot, ensuring the feasibility of the final output architectural design model on the architectural design plot, effectively avoiding unreasonable design schemes, and improving practicality and feasibility.

[0009] Optionally, the architectural design modeling system further comprises the following steps: B1, generating corresponding plot area model data according to the design plot area surveying and mapping data; B2, obtaining the plot model positioning data corresponding to the building design plot in the plot area model data; B3, adding each available building design model data to the location data of the plot model to generate corresponding model data of each building design area; B4, obtaining the corresponding regional building top view data according to the model data of each building design area; B5, identifying and generating corresponding road path data using a preset road recognition algorithm according to the regional building bird's-eye view data; B6, determining corresponding adjacent road path data according to the land parcel model positioning data and the road path data; B7, according to the adjacent road path data and the plot model positioning data, a preset virtual camera is used to obtain virtual appearance image data of each designed building corresponding to the available building design model data in the building design area model data.

[0010] By adopting the above technical solution, the building design modeling system can automatically generate building appearance images from multiple perspectives by placing the building model in the surveying and mapping model of the design plot and analyzing the surrounding traffic environment in combination with the road recognition algorithm. This can not only help designers more intuitively evaluate the integration effect of the building and the surrounding environment, but also provide important visual reference basis for subsequent design optimization, thereby improving the visualization level of building design and the environmental adaptability assessment capability.

[0011] Optionally, the architectural design modeling system further comprises the following steps: B8, generating corresponding design building rendering data through a preset image beautification model according to the virtual appearance image data of each design building corresponding to the architectural design area model data; B9, generating corresponding rendering feature vectors according to all the design building rendering data using a preset feature extraction algorithm; B10, calculating the corresponding effect graph feature center vector according to each effect graph feature vector; B11, calculating the corresponding cosine similarity of the effect graph according to the feature vector of each effect graph and the feature center vector of the effect graph; B12, if the cosine similarity of the effect picture corresponding to the effect picture feature vector is greater than a preset effect picture similarity threshold, then define the design building effect picture data corresponding to the effect picture feature vector as the classified effect picture data; B13, generating design building renderings set data according to the combination of all classified renderings data.

[0012] By adopting the above technical solution, the architectural design modeling system can optimize the original renderings through the image beautification model, and use feature vector analysis and cosine similarity calculation to evaluate and screen the renderings, ensuring that the final generated rendering collection has high visual quality and style consistency. It can not only ensure the professionalism and aesthetics of the renderings, but also automatically eliminate pictures with large style differences, providing designers with high-quality and unified style architectural rendering collections.

[0013] Optionally, the architectural design modeling system further comprises the following steps: B14, determining the corresponding design building rendering data according to the rendering feature vector with the largest cosine similarity of the renderings and defining it as typical building rendering data; B15, generating corresponding typical rendering description text through the image description module according to the typical building rendering data; B16, generating corresponding design building effect graphic and text data according to the design building effect atlas data and the typical effect description text combination.

[0014] By adopting the above technical solution, the architectural design modeling system can select the most representative design building renderings and automatically generate corresponding text descriptions, which can combine visual presentation and text description to provide a more comprehensive and professional display method for the design scheme. It can not only help relevant personnel better understand the design intent, but also provide a more professional and standardized design results display plan, thereby improving the expression effect and communication efficiency of architectural design results.

[0015] Optionally, the architectural design modeling system further comprises the following steps: C1, respectively set the corresponding simulated daylight in the model data of each building design area; C2, continuously adjusting the solar altitude angle and solar azimuth angle of the simulated sunlight according to the positioning data of the building design plot and the preset simulation time window; C3, obtaining a corresponding simulation time when adjusting the simulated daylight, and continuously obtaining simulated lighting image data corresponding to the available architectural design model data in the architectural design area model data at a preset lighting shooting position; C4, calculating the simulated lighting data corresponding to the available architectural design model data in the architectural design area model data according to the simulated lighting image data corresponding to all simulation moments.

[0016] By adopting the above technical solution, the building design modeling system can simulate the sunlight conditions at different times. The system can automatically analyze the lighting effect of the building under all-weather conditions and evaluate the building's natural light utilization by continuously collecting lighting data. It can not only help designers to more accurately evaluate the building's lighting performance, but also provide important data support for building energy-saving design, effectively improve the scientificity and sustainability of building design, and provide a data path for subsequent improvements in building design.

[0017] Optionally, the architectural design modeling system further comprises the following steps: C5, comprehensively calculating and determining the corresponding average daylighting duration according to the simulated daylighting data corresponding to the available building design model data in the model data of each building design area; C6, if the average daylighting duration is less than the preset daylighting duration threshold, return to step A5 and add the corresponding daylighting duration description text to the architectural design requirement text.

[0018] By adopting the above technical solution, the building design modeling system can calculate the average daylighting time and compare it with the preset standards. When the daylighting conditions do not meet the requirements, it will automatically feed back the daylighting problem to the design concept stage and add relevant descriptions in the design requirements, thereby prompting the system to generate new design solutions with better lighting performance. This closed-loop optimization mechanism ensures that the building design can improve the situation of insufficient daylighting, which can effectively improve the practicality and living comfort of the design solution.

[0019] Optionally, step A6 includes the following steps: A601, generating corresponding building pre-drawing description text through the image description module according to each building pre-drawing data; A602, generating a corresponding pre-construction description text word feature vector according to the architectural pre-construction description text by using a preset word vector algorithm; A603, generating a corresponding concept text word feature vector according to the architectural design concept text by using a word vector algorithm; A604, calculating the corresponding text cosine similarity according to the pre-constructed description text word feature vector and the conception text word feature vector; A605, if the text cosine similarity is greater than a preset text similarity threshold, defining the architectural pre-image data corresponding to the pre-image description text word feature vector as valid architectural pre-image data; A606, generating a pre-constructed building image dataset according to all valid pre-constructed building image data combinations.

[0020] By adopting the above technical solution, the architectural design modeling system can convert the architectural pre-drawing into a text description and use the word vector algorithm to calculate the similarity with the design concept, so as to screen out the pre-drawing scheme that matches the design intent. It can not only ensure that the generated architectural scheme is consistent with the design concept, but also effectively filter out the scheme that does not meet the design intent, thereby improving the accuracy of the design results and the design efficiency.

[0021] In summary, the present application includes at least one of the following beneficial technical effects: 1. By combining surveying and mapping data, image analysis, text description and model generation, it is possible to consider both the characteristics of the plot and the surrounding environmental factors at the same time, thereby generating a building design solution that better meets the actual scene requirements. By generating a prefabricated building image data set, multiple design solutions are provided for selection, which increases the flexibility and selectivity of the design. The entire process is highly automated, which can significantly improve the efficiency of building design while ensuring the coordination of the design solution with the surrounding environment.

[0022] 2. The matching degree analysis can be performed between the building bottom surface image of the generated building prefabricated model data and the actual plot image to increase the model verification and screening mechanism, evaluate the adaptability of the building design scheme and the plot, ensure the feasibility of the final output building design model on the building design plot, effectively avoid unreasonable design schemes, and improve practicality and feasibility.

[0023] 3. By placing the building model in the surveying and mapping model of the design site and combining it with the road recognition algorithm to analyze the surrounding traffic environment, the building appearance image can be automatically generated from multiple perspectives. This can not only help designers more intuitively evaluate the integration effect of the building and the surrounding environment, but also provide important visual reference for subsequent design optimization, thereby improving the visualization level of architectural design and the ability to evaluate environmental adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a principle schematic diagram of a building design modeling system of the present invention.

[0025] Figure 2 It is a process schematic diagram of an architectural design modeling strategy of an architectural design modeling system of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0027] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0028] refer to Figure 1 The present invention provides a building design modeling system for performing building design based on an AI model, comprising: Mapping module 10; Image description module 20; Image generation module 30; Model generation module 40; Processing control module 50; The surveying and mapping module 10 , the image description module 20 , the image generation module 30 , and the model generation module 40 are respectively data-connected to the processing control module 50 .

[0029] The surveying and mapping module 10 is mainly used to survey the terrain and building data of the corresponding land plot and its surrounding environment that require architectural design. The surveying and mapping module 10 can be a surveying and mapping module based on a drone.

[0030] The image description module 20 is mainly used to input an image and generate a description text content of the corresponding image.

[0031] The image generation module 30 is mainly used to generate corresponding image content according to the input text content.

[0032] The model generation module 40 is mainly used to generate corresponding three-dimensional model data according to the input image content.

[0033] The processing control module 50 is mainly used for data processing and controlling other modules.

[0034] Wherein, the architectural design modeling system further includes an architectural design modeling strategy, including the following steps: A1, obtaining corresponding design plot area surveying data through surveying by the surveying module 10 according to the preset building design plot and the preset surveying radius; The architectural design plot is the construction plot that needs to be designed; The surveying radius is a preset radius length, which is used to determine the corresponding surveying range with the building design plot as the center; The design plot area surveying and mapping data refers to the terrain and building surveying and mapping data within the surveying and mapping radius area with the building design plot as the center, that is, the terrain and building surveying and mapping data within the surveying and mapping radius around the building design plot.

[0035] A2, extracting corresponding land parcel top view image data and land parcel surrounding scene image data from the designed land parcel area surveying and mapping data through the processing control module 50; The plot overhead image data is a overhead image of the building design plot; The plot surrounding scene image data is image data of the terrain or buildings around the building design plot.

[0036] A3, generating corresponding land parcel feature description text through the image description module 20 according to the land parcel overhead image data; The plot feature description text is the descriptive text content corresponding to the plot overhead image data, which can be a qualitative description of the size and shape of the building design plot in the plot overhead image data. It can be generated by inputting the plot overhead image data into the image description module 20. There are many existing multimodal large models that can generate corresponding description texts based on the input image.

[0037] A4, generating corresponding surrounding scene feature description text through the image description module 20 according to the surrounding scene image data of the land parcel; The surrounding scene feature description text is the descriptive text content corresponding to the scene image data surrounding the plot, and can be a qualitative description of the environmental conditions, building style, building height and building density around the building design plot.

[0038] A5, generating a corresponding architectural design concept text by combining the plot feature description text, the surrounding scene feature description text and the preset architectural design requirement text through the processing control module 50; The architectural design requirements are the requirements prepared in advance by the designers, such as design style, architectural color, building height and ancillary facilities; The architectural design concept text is a collection of block feature description text, surrounding scene feature description text and architectural design requirement text, which is used to comprehensively reflect the actual situation and design requirements of the building.

[0039] A6, generating corresponding pre-constructed image data of each building through the image generation module 30 according to the architectural design concept text and the preset number of image generation, and combining the pre-constructed image data of each building to generate a pre-constructed image data set of the building; The number of images generated is a preset number value, which is used to generate a corresponding number of architectural pre-composition data through the image generation module 30 according to the architectural design concept text; The architectural pre-drawing data is an architectural design drawing generated by the image generation module 30 according to the architectural design concept text, and the corresponding architectural image can be generated by inputting the design requirements in the form of text through the existing multi-modal large model; The architectural pre-constructed image dataset is a dataset of all architectural pre-constructed image data; Since there is a certain degree of randomness in the image content generated by the multimodal large model, setting the number of images generated can ensure that there is sufficient architectural pre-composition data for subsequent data processing and screening.

[0040] A7, generating corresponding prefabricated building model data through the model generation module 40 for each prefabricated building image data in the prefabricated building image data set; The building prefabricated model data is the building three-dimensional model data generated after the building prefabricated drawing data is input into the model generation module 40 .

[0041] Through the above technical solutions, the architectural design modeling system can combine surveying and mapping data, image analysis, text description and model generation, and can simultaneously consider the characteristics of the plot and the surrounding environmental factors, thereby generating an architectural design solution that better meets the actual scene requirements. Through the generation of a prefabricated architectural image data set, multiple design solutions are provided for selection, which increases the flexibility and selectivity of the design. The entire process is highly automated, which can significantly improve the efficiency of architectural design while ensuring the coordination of the design solution with the surrounding environment.

[0042] Furthermore, the architectural design modeling system further comprises the following steps: A8, determining corresponding building model bottom surface image data according to each building prefabricated model data; The building model bottom surface image data is a bottom surface image of a three-dimensional model corresponding to the building prefabricated model data, and is used to reflect the shape of the area occupied by the building prefabricated model data.

[0043] A9, calculating the graphic matching degree of the corresponding building plot according to the bottom surface image data of the building model and the top view image data of the plot using a preset graphic matching algorithm; The graphic matching algorithm calculation is a pre-set algorithm used to calculate the matching degree between two images. There are many related algorithms that can realize the calculation of graphic matching degree, such as gray-scale-based matching algorithm, feature-based matching algorithm and relationship-based matching algorithm, which can be selected and used according to needs; The building plot graphic matching degree is the matching degree value between the bottom surface image data of the building model and the top view image data of the plot. If the building plot graphic matching degree is low, it means that the building prefabricated model data corresponding to the bottom surface image data of the building model may not match the building design plot, and thus construction cannot be carried out.

[0044] A10, if the matching degree of the building plot graphic is not less than a preset matching threshold, the corresponding prefabricated building model data is defined as the available building design model data; The matching threshold is a pre-set reference value used to determine whether the matching degree of the building plot graphics meets the requirements; The available architectural design model data is prefabricated architectural model data that is determined to be estimated to be able to be constructed on the architectural design plot.

[0045] Through the above steps, the architectural design modeling system can perform a matching degree analysis between the building bottom surface image of the generated building prefabricated model data and the actual plot image to increase the model verification and screening mechanism, and can evaluate the adaptability of the architectural design scheme and the plot, ensure the feasibility of the final output architectural design model on the architectural design plot, effectively avoid unreasonable design schemes, and improve practicality and feasibility.

[0046] Furthermore, the architectural design modeling system further comprises the following steps: B1, generating corresponding plot area model data according to the design plot area surveying and mapping data; The plot area model data is three-dimensional model data generated based on the design plot area surveying and mapping data, that is, the three-dimensional model data of the area environment and buildings around the building design plot.

[0047] B2, obtaining the plot model positioning data corresponding to the building design plot in the plot area model data; The plot model positioning data is the positioning data of the building design plot in the plot area model data.

[0048] B3, adding each available building design model data to the location data of the plot model to generate corresponding model data of each building design area; The architectural design area model data is three-dimensional model data generated by adding available architectural design model data to the architectural design plot in the plot area model data, that is, the available architectural design model data model is virtually built in the plot area model data.

[0049] B4, obtaining the corresponding regional building top view data according to the model data of each building design area; The regional building bird's-eye view data is a bird's-eye view corresponding to the architectural design regional model data.

[0050] B5, identifying and generating corresponding road path data using a preset road recognition algorithm according to the regional building bird's-eye view data; The road recognition algorithm is a pre-set algorithm used to identify the location and direction data of the corresponding road based on the bird's-eye view; The road path data is the road location and direction data in the regional building overhead view data.

[0051] B6, determining corresponding adjacent road path data according to the land parcel model positioning data and the road path data; The adjacent road path data are location and direction data of roads adjacent to the available architectural design model data in the architectural design area model data.

[0052] B7, acquiring virtual appearance image data of each designed building corresponding to the available building design model data by using a preset virtual camera in the building design area model data according to the adjacent road path data and the plot model positioning data; The virtual camera is a virtual visual point set in the three-dimensional model for image acquisition, including corresponding position and orientation angle data; By moving the virtual camera according to the adjacent road path data and keeping the virtual camera always facing the available building design model data in the building design area model data according to the plot model positioning data, the appearance images corresponding to the available building design model data can be obtained from multiple directions and angles based on the vision of pedestrians.

[0053] The virtual appearance image data of the designed building can be the appearance image corresponding to the architectural design model data in the architectural design area model data, which can be used for reference and evaluation by designers or for subsequent data analysis and processing.

[0054] Through the above steps, the architectural design modeling system can automatically generate building appearance images from multiple perspectives by placing the building model in the surveying and mapping model of the design plot and analyzing the surrounding traffic environment in combination with the road recognition algorithm. This can not only help designers more intuitively evaluate the integration effect of the building and the surrounding environment, but also provide important visual reference for subsequent design optimization, thereby improving the visualization level of architectural design and the environmental adaptability assessment capability.

[0055] Furthermore, the architectural design modeling system further comprises the following steps: B8, generating corresponding design building rendering data through a preset image beautification model according to the virtual appearance image data of each design building corresponding to the architectural design area model data; The image beautification model is a pre-trained or pre-selected model used to beautify and enhance the input image. It can be a dedicated image beautification model or an existing multi-modal large model. The design building rendering data is rendering data generated after beautifying the design building virtual appearance image data.

[0056] B9, generating corresponding rendering feature vectors according to all the design building rendering data using a preset feature extraction algorithm; The rendering feature vector is a feature vector corresponding to the design building rendering data.

[0057] B10, calculating the corresponding effect graph feature center vector according to each effect graph feature vector; The effect graph feature center vector is the center vector of all effect graph feature vectors. The center vector can be determined by clustering algorithm calculation or by averaging the dimensions of the effect graph feature vectors.

[0058] B11, calculating the corresponding cosine similarity of the effect graph according to the feature vector of each effect graph and the feature center vector of the effect graph; The cosine similarity of the effect image is the cosine value of the angle between the effect image feature vector and the effect image feature center vector.

[0059] B12, if the cosine similarity of the effect picture corresponding to the effect picture feature vector is greater than a preset effect picture similarity threshold, then define the design building effect picture data corresponding to the effect picture feature vector as the classified effect picture data; The effect graph similarity threshold is a pre-set reference value used to determine whether the similarity between the effect graph feature vector and the effect graph feature center vector meets the requirement; The classified rendering data is a data collection of design and building rendering data whose cosine similarity of the renderings meets the rendering similarity threshold, which can be understood as design and building rendering data that can be classified with similar design styles.

[0060] B13, generating design building effect drawing set data according to all the classified effect drawing data combinations; The design building renderings collection data is a collection of all classified renderings data.

[0061] Through the above steps, the architectural design modeling system can optimize the original renderings through the image beautification model, and use feature vector analysis and cosine similarity calculation to evaluate and screen the renderings, ensuring that the final generated renderings have high visual quality and style consistency. It can not only ensure the professionalism and aesthetics of the renderings, but also automatically eliminate pictures with large style differences, providing designers with high-quality and unified style architectural renderings.

[0062] Furthermore, the architectural design modeling system further comprises the following steps: B14, determining the corresponding design building rendering data according to the rendering feature vector with the largest cosine similarity of the renderings and defining it as typical building rendering data; The typical building rendering data is the design building rendering data of the rendering feature vector with the largest rendering cosine similarity, that is, the most representative design building rendering data in the design building rendering set data.

[0063] B15, generating corresponding typical rendering description text through the image description module 20 according to the typical architectural rendering data; The typical rendering description text is the descriptive text content corresponding to the typical building rendering data.

[0064] B16, generating corresponding design building effect graphic and text data according to the design building effect atlas data and the typical effect description text combination; The design building rendering graphic data is a collection of design building rendering atlas data and typical rendering description texts.

[0065] Through the above steps, the architectural design modeling system can select the most representative design building renderings and automatically generate corresponding text descriptions, which can combine visual presentation and text description to provide a more comprehensive and professional display method for design plans. It can not only help relevant personnel better understand the design intent, but also provide more professional and standardized design results display plans, thereby improving the expression effect and communication efficiency of architectural design results.

[0066] Furthermore, the architectural design modeling system further comprises the following steps: C1, respectively set the corresponding simulated daylight in the model data of each building design area; Simulated daylight can be set up through corresponding 3D software to simulate the movement of the sun, and the corresponding altitude angle and azimuth angle can be calculated according to the algorithm of the sun's movement; By setting simulated daylight in the architectural design area model data, the lighting effects of adjacent buildings and environmental terrain on the available architectural design model data can be fully considered.

[0067] C2, continuously adjusting the solar altitude angle and solar azimuth angle of the simulated sunlight according to the positioning data of the building design plot and the preset simulation time window; The simulation time window is a pre-set time window used to determine the time range for the operation of simulated sunlight; The solar altitude angle is the altitude angle of the sun simulated by the simulated sunlight in the architectural design area model data; The solar azimuth is the azimuth of the sun simulated in the architectural design area model data by simulating sunlight.

[0068] C3, obtaining a corresponding simulation time when adjusting the simulated daylight, and continuously obtaining simulated lighting image data corresponding to the available architectural design model data in the architectural design area model data at a preset lighting shooting position; The simulation time is the corresponding time in the simulation time window when the simulated daylight operation is adjusted; The lighting shooting position is a pre-set virtual shooting position, which is used to obtain the lighting image of the available architectural design model data in the architectural design area model data. For example, a corresponding position can be set in the south, east, west or north of the available architectural design model data to continuously acquire images facing the available architectural design model data, so as to obtain different lighting images of the available architectural design model data at different simulation times; The simulated lighting image data is the lighting image images at different simulation times of the available building design model data.

[0069] C4, calculating the simulated lighting data corresponding to the available architectural design model data in the architectural design area model data according to the simulated lighting image data corresponding to all simulation moments; The simulated lighting data is data on the lighting conditions at a given moment within a simulated time window using the available architectural design model data, and can be determined by calculating the simulated lighting image data. For example, the simulated lighting data can be determined by calculating the corresponding illuminated area based on the brightness data of the simulated lighting image data.

[0070] Through the above steps, the building design modeling system can simulate the sunlight conditions at different times. The system can automatically analyze the lighting effect of the building under all-weather conditions and evaluate the building's natural light utilization by continuously collecting lighting data. It can not only help designers to more accurately evaluate the building's lighting performance, but also provide important data support for building energy-saving design, effectively improve the scientificity and sustainability of building design, and provide a data path for subsequent improvements in building design.

[0071] Furthermore, the architectural design modeling system further comprises the following steps: C5, comprehensively calculating and determining the corresponding average daylighting duration according to the simulated daylighting data corresponding to the available building design model data in the model data of each building design area; The average daylighting duration is the average value of the daylighting duration determined by calculating the simulated daylighting data, which can be the overall average duration or the average duration of a certain season.

[0072] C6, if the average daylighting duration is less than the preset daylighting duration threshold, return to step A5 and add the corresponding daylighting duration description text to the architectural design requirement text.

[0073] The daylighting duration threshold is a pre-set reference value used to determine whether the average daylighting duration meets the set requirements; If the average daylighting duration is insufficient, by returning to step A5 and adding a daylighting duration description text with contents such as insufficient daylighting to the architectural design requirement text, the daylighting condition can be fully considered when the architectural pre-composition data is generated again.

[0074] Through the above steps, the architectural design modeling system can calculate the average daylighting time and compare it with the preset standards. When the daylighting conditions do not meet the requirements, it will automatically feed back the daylighting problem to the design concept stage and add relevant descriptions in the design requirements, thereby prompting the system to generate new design solutions with better lighting performance. This closed-loop optimization mechanism ensures that the architectural design can improve the situation of insufficient daylighting, which can effectively improve the practicality and living comfort of the design solution.

[0075] Furthermore, the step A6 comprises the following steps: A601, generating corresponding building pre-drawing description text through the image description module 20 according to each building pre-drawing data; The architectural pre-drawing description text is the descriptive text content corresponding to the architectural pre-drawing data.

[0076] A602, generating a corresponding pre-construction description text word feature vector according to the architectural pre-construction description text by using a preset word vector algorithm; The word vector algorithm is a pre-set algorithm used to generate corresponding word feature vectors based on the text content; The word feature vector of the pre-construction description text is the word feature vector corresponding to the building pre-construction description text.

[0077] A603, generating a corresponding concept text word feature vector according to the architectural design concept text by using a word vector algorithm; The concept text word feature vector is a word feature vector corresponding to the architectural design concept text.

[0078] A604, calculating the corresponding text cosine similarity according to the pre-constructed description text word feature vector and the conception text word feature vector; The text cosine similarity is the cosine similarity between the pre-constructed description text word feature vector and the concept text word feature vector.

[0079] A605, if the text cosine similarity is greater than a preset text similarity threshold, defining the architectural pre-image data corresponding to the pre-image description text word feature vector as valid architectural pre-image data; The text similarity threshold is a pre-set reference value used to determine whether the text cosine similarity reaches a corresponding similarity level; The valid architectural pre-composition data are architectural pre-composition data whose text cosine similarity meets the requirements, that is, the corresponding architectural pre-composition data meet the requirements in the architectural design concept text to a certain extent.

[0080] A606, generating a pre-constructed building image data set according to all valid pre-constructed building image data combinations; The architectural pre-construction image dataset is a collection of all valid architectural pre-construction image data.

[0081] Through the above steps, the architectural design modeling system can convert the architectural pre-drawing into a text description and use the word vector algorithm to calculate the similarity with the design concept, so as to screen out the pre-drawing scheme that matches the design intent. It can not only ensure that the generated architectural scheme is consistent with the design concept, but also effectively filter out the scheme that does not meet the design intent, thereby improving the accuracy of the design results and the design efficiency.

[0082] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. A building design modeling system, characterized in that: include: Surveying and mapping module; Image description module; Image generation module; Model generation module; Processing control module; Wherein, the surveying and mapping module, the image description module, the image generation module, and the model generation module are respectively data-connected to the processing control module; Wherein, the architectural design modeling system further includes an architectural design modeling strategy, including the following steps: A1, obtaining corresponding design plot area surveying data through surveying by the surveying module according to the preset building design plot and the preset surveying radius; A2, extracting corresponding land parcel top view image data and land parcel surrounding scene image data from the designed land parcel area surveying and mapping data through the processing control module; A3, generating corresponding land parcel feature description text through the image description module according to the land parcel overhead image data; A4, generating corresponding surrounding scene feature description text through the image description module according to the surrounding scene image data of the land parcel; A5, generating a corresponding architectural design concept text by combining the plot feature description text, the surrounding scene feature description text and the preset architectural design requirement text through the processing control module; A6, generating corresponding pre-constructed image data of each building through the image generation module according to the architectural design concept text and the preset number of image generation, and combining the pre-constructed image data of each building to generate a pre-constructed image data set of the building; A7, generating corresponding prefabricated building model data through the model generation module for each prefabricated building image data in the prefabricated building image data set.

2. The architectural design modeling system according to claim 1, characterized in that: Further comprising the steps of: A8, determining corresponding building model bottom surface image data according to each building prefabricated model data; A9, calculating the graphic matching degree of the corresponding building plot according to the bottom surface image data of the building model and the top view image data of the plot using a preset graphic matching algorithm; A10: If the matching degree of the building plot graphic is not less than a preset matching threshold, the corresponding prefabricated building model data is defined as available building design model data.

3. The architectural design modeling system according to claim 2, characterized in that: Further comprising the steps of: B1, generating corresponding plot area model data according to the design plot area surveying and mapping data; B2, obtaining the plot model positioning data corresponding to the building design plot in the plot area model data; B3, adding each available building design model data to the location data of the plot model to generate corresponding model data of each building design area; B4, obtaining the corresponding regional building top view data according to the model data of each building design area; B5, identifying and generating corresponding road path data using a preset road recognition algorithm according to the regional building bird's-eye view data; B6, determining corresponding adjacent road path data according to the land parcel model positioning data and the road path data; B7, according to the adjacent road path data and the plot model positioning data, a preset virtual camera is used to obtain virtual appearance image data of each designed building corresponding to the available building design model data in the building design area model data.

4. The architectural design modeling system according to claim 3, characterized in that: Further comprising the steps of: B8, generating corresponding design building rendering data through a preset image beautification model according to the virtual appearance image data of each design building corresponding to the architectural design area model data; B9, generating corresponding rendering feature vectors according to all the design building rendering data using a preset feature extraction algorithm; B10, calculating the corresponding effect graph feature center vector according to each effect graph feature vector; B11, calculating the corresponding cosine similarity of the effect graph according to the feature vector of each effect graph and the feature center vector of the effect graph; B12, if the cosine similarity of the effect picture corresponding to the effect picture feature vector is greater than a preset effect picture similarity threshold, then define the design building effect picture data corresponding to the effect picture feature vector as the classified effect picture data; B13, generating design building renderings set data according to the combination of all classified renderings data.

5. The architectural design modeling system according to claim 4, characterized in that: Further comprising the steps of: B14, determining the corresponding design building rendering data according to the rendering feature vector with the largest cosine similarity of the renderings and defining it as typical building rendering data; B15, generating corresponding typical rendering description text through the image description module according to the typical building rendering data; B16, generating corresponding design building effect graphic and text data according to the design building effect atlas data and the typical effect description text combination.

6. The architectural design modeling system according to claim 5, characterized in that: Further comprising the steps of: C1, respectively set the corresponding simulated daylight in the model data of each building design area; C2, continuously adjusting the solar altitude angle and solar azimuth angle of the simulated sunlight according to the positioning data of the building design plot and the preset simulation time window; C3, obtaining a corresponding simulation time when adjusting the simulated daylight, and continuously obtaining simulated lighting image data corresponding to the available architectural design model data in the architectural design area model data at a preset lighting shooting position; C4, calculating the simulated lighting data corresponding to the available architectural design model data in the architectural design area model data according to the simulated lighting image data corresponding to all simulation moments.

7. The architectural design modeling system according to claim 6, characterized in that: Further comprising the steps of: C5, comprehensively calculating and determining the corresponding average daylighting duration according to the simulated daylighting data corresponding to the available building design model data in the model data of each building design area; C6, if the average daylighting duration is less than the preset daylighting duration threshold, return to step A5 and add the corresponding daylighting duration description text to the architectural design requirement text.

8. The architectural design modeling system according to claim 7, characterized in that: The step A6 comprises the following steps: A601, generating corresponding building pre-drawing description text through the image description module according to each building pre-drawing data; A602, generating a corresponding pre-construction description text word feature vector according to the architectural pre-construction description text by using a preset word vector algorithm; A603, generating a corresponding concept text word feature vector according to the architectural design concept text by using a word vector algorithm; A604, calculating the corresponding text cosine similarity according to the pre-constructed description text word feature vector and the conception text word feature vector; A605, if the text cosine similarity is greater than a preset text similarity threshold, defining the architectural pre-image data corresponding to the pre-image description text word feature vector as valid architectural pre-image data; A606, generating a pre-constructed building image dataset according to all valid pre-constructed building image data combinations.

Images