A Novel Intelligent Generation Method for the Skeleton Structure of Yurts Based on Deep Learning

Abstract

The present invention discloses a novel intelligent generation method for the framework structure of a yurt based on deep learning, comprising the following steps: three-dimensional model data acquisition of the yurt framework structure, topological transformation of the yurt framework structure and prefabricated components, construction of a framework structure generation model based on a graph neural network, generation of the yurt framework structure model and matching generation of prefabricated components, stability testing of the yurt framework structure model, and output of component drawings. The advantages are as follows: it realizes the automatic extraction of the topological relationships among the Hana components, Uni components, and Taonao components in the framework structure and among the individual components that make them up, reflects the connection relationships among the components, reduces the workload of identifying the connection relationships of the components, and improves the work efficiency of architects; it improves the intelligence and automation level of the yurt framework structure, enhances the accuracy of generating standardized prefabricated components, and reduces the difficulty of procurement, production, and assembly of prefabricated components.

Description

Technical Field:

[0001] The present invention belongs to the technical field of building styles, and particularly relates to a method for intelligently generating a new type of yurt skeleton structure based on deep learning. Background Art:

[0002] The intelligent generation of building styles is one of the important contents of intelligent construction and an important means to improve the efficiency of building design. The application of artificial intelligence technology can not only learn from a large number of case samples but also generate multiple design schemes for architects to compare and select within a short time, accelerating the design process. Currently, obtaining and organizing a large amount of building style data and conducting effective learning and generation are important challenges in building style generation. The yurt is a special type of building, which is very different from modern buildings and traditional Chinese wooden buildings in terms of component materials, component styles, and skeleton structure systems. Traditional manual design methods require a lot of time and effort, and currently, there is also a lack of an intelligent generation method for yurt building styles, which cannot meet the need for efficiently designing yurt buildings. Summary of the Invention:

[0003] Aiming at the deficiencies of the prior art, the present invention provides a method for intelligently generating a new type of yurt skeleton structure based on deep learning, which solves the problem of time-consuming and laborious traditional Mongolian building design and provides a solution for the intelligent generation of Mongolian building styles.

[0004] The present invention is implemented by the following technical solutions: A method for intelligently generating a new type of yurt skeleton structure based on deep learning, the method comprising the following steps:

[0005] Step (1), collecting three-dimensional model data of the yurt building skeleton structure,

[0006] Using a laser scanning device to measure the objects of the yurt skeleton, collecting the point cloud data of the yurt skeleton structure, classifying the point cloud data through a point cloud classification algorithm to obtain the point cloud data of the khana component, the tolno component, and the uni component, inputting the point cloud data into three-dimensional modeling software to generate a three-dimensional model of the yurt skeleton structure composed of khana, tolno, and uni, and constructing a three-dimensional model library of the yurt skeleton structure;

[0007] Step (2), topological conversion of the Mongolian building skeleton structure and precast components,

[0008] Calculate the three-dimensional morphological attributes of the Hana component, Taonao component, and Uni component in step (1), including length, width, height, and cross-sectional shape, and construct a component three-dimensional morphology database; identify and extract the positions where each component is connected as connection points, and according to the connection relationship between the component and the connection point, transform the three-dimensional data of the skeleton structure into topological data, and connect the connection point with the corresponding component data in the component three-dimensional morphology database to establish a sample topological data set, where the training data set accounts for 80% and the test data set accounts for 20%;

[0009] Step (3), construction of a skeleton structure generation model based on a graph neural network

[0010] Input the training data set in step (2) into a deep learning workstation equipped with a graph neural network deep learning model, extract and learn the rules of the structural skeleton topology network, and construct a ger skeleton structure generation model;

[0011] (4) Generation of Mongolian architectural structure styles and matching generation of prefabricated components

[0012] According to the basic structure data of the ger, establish a ger basic model database, construct a ger skeleton structure generation system, input the area and height data of the ger to be built into the system, and the ger skeleton structure generation system can automatically match the basic model with the smallest error from the input area and height data, and input the basic model into the deep learning workstation in step (3) to automatically generate a skeleton structure model, and provide the data information and required quantity information of the three-dimensional models of the standardized prefabricated components that make up the Hana component, Taonao component, and Uni component;

[0013] Step (5), skeleton structure stability test

[0014] Input the skeleton structure model generated in step (4) into engineering simulation software for structural mechanics analysis. For the skeleton structure that generates structural displacement and deformation, return to step (4) to regenerate the skeleton structure model;

[0015] (6) Output of component drawings

[0016] Output the list and drawings of the required prefabricated components as excel and cad format files for use in the next procurement and construction.

[0017] Furthermore, the point cloud classification algorithm in step (1) is a deep learning algorithm trained by a point cloud data set, which can automatically classify the unprocessed point cloud data.

[0018] Furthermore, the identification and extraction of the positions where each component is connected as connection points in step (2) refers to identifying by calculating the distance between the surfaces of the components, and the position where the distance is 0 is identified as the connection point.

[0019] Further, the transformation of the three-dimensional data of the skeleton structure into topological data in step (2) means taking the components in the skeleton structure as the points in the topological network and the connecting points as the edges of the connection points.

[0020] Further, the extraction and learning of the rules of the structural skeleton topological network in step (3) include the connection relationship between Hana components and the number of connection points, the connection relationship between Hana components and Uni components, the connection relationship between Uni components and Taonao components and the number of connection points, and the connection relationship between Taonao components and the number of connection points.

[0021] Further, the basic structure data of the yurt in step (4) includes the bottom diameter of the yurt, the height of the Hana wall, the diameter of the Taonao, the height of the bottom edge of the Taonao, and the total height of the yurt.

[0022] Further, the basic model database of the yurt in step (4) is a database of the yurt skeleton structure model storing the basic structure data of various yurts.

[0023] Further, the structural mechanics analysis in step (5) means analyzing the stress, strain and deformation of the structural skeleton under static load through engineering simulation software such as ANSYS.

[0024] Further, it is characterized in that the list and drawings of the prefabricated components in step (6) include the data of the type, quantity, diameter and length of the prefabricated components.

[0025] Advantages of the present invention: Compared with the prior art, the beneficial effects of the present invention are as follows:

[0026] By performing topological conversion on the skeleton structure of the yurt, the automatic extraction of the topological relationships between Hana components, Uni components, Taonao components and between the individual components constituting them in the skeleton structure is realized, reflecting the connection relationships between the components, reducing the workload of identifying the connection relationships between the components, and improving the work efficiency of architects;

[0027] By applying the graph neural network deep learning model, it is possible to automatically learn the topological data of a large number of yurt skeleton structures, improving the efficiency of topological data learning, and being able to more comprehensively understand the connection relationships of the yurt skeleton components compared with other machine learning algorithms, improving the accuracy of extracting and learning the rules of the topological network structure;

[0028] By generating the skeleton structure model and the information of standardized prefabricated components through the yurt skeleton structure generation system based on the graph neural network, the intelligence and automation level of the yurt skeleton structure are improved, the accuracy of generating standardized prefabricated components is improved, and the difficulty of purchasing, producing and assembling prefabricated components is reduced. Description of the drawings:

[0029] Figure 1 is the flowchart of the present invention. Specific implementation mode:

[0030] Example 1: A novel intelligent generation method for the framework structure of a yurt based on deep learning, the method comprising the following steps:

[0031] Step (1), three-dimensional model data acquisition of the framework structure of the yurt building,

[0032] Use a laser scanning device to measure the object of the yurt framework, collect the point cloud data of the yurt framework structure, classify the point cloud data through a point cloud classification algorithm, obtain the point cloud data of the khana component, the tolno component, and the uni component, input the point cloud data into three-dimensional modeling software, generate a three-dimensional model of the yurt framework structure composed of khana, tolno, and uni, and construct a three-dimensional model library of the yurt framework structure;

[0033] Step (2), topological conversion of the framework structure and prefabricated components of Mongolian architecture,

[0034] Calculate the three-dimensional morphological attributes of the khana component, the tolno component, and the uni component in step (1), including length, width, height, and cross-sectional shape, and construct a three-dimensional morphological database of components; identify and extract the positions where each component is connected as connection points, and according to the connection relationship between the component and the connection point, convert the three-dimensional data of the framework structure into topological data, and connect the connection point with the corresponding component data in the three-dimensional morphological database of components to establish a sample topological data set, where the training data set accounts for 80% and the test data set accounts for 20%;

[0035] Step (3), construction of a framework structure generation model based on a graph neural network,

[0036] Input the training data set in step (2) into a deep learning workstation equipped with a graph neural network deep learning model, extract and learn the rules of the structure framework topology network, and construct a yurt framework structure generation model;

[0037] (4) Generation of the structure style of Mongolian architecture and matching generation of prefabricated components,

[0038] According to the basic structure data of the yurt, establish a basic model database of the yurt, construct a yurt framework structure generation system, input the area and height data of the yurt to be built into the system, the yurt framework structure generation system can automatically match the basic model with the smallest error from the input area and height data, and input the basic model into the deep learning workstation in step (3) to automatically generate a framework structure model, and provide the data information and required quantity information of the standardized prefabricated component three-dimensional models that make up the khana component, the tolno component, and the uni component;

[0039] Step (5), stability test of the skeleton structure

[0040] Input the skeleton structure model generated in step (4) into engineering simulation software for structural mechanics analysis. For the skeleton structure with structural displacement and deformation, return to step (4) to regenerate the skeleton structure model;

[0041] (6) Output of component drawings

[0042] Output the list and drawings of prefabricated components required as excel and cad format files for the next procurement and construction use.

[0043] Furthermore, the point cloud classification algorithm in step (1) is a deep learning algorithm trained by a point cloud data set, which can automatically classify the unprocessed point cloud data.

[0044] Furthermore, the step of identifying and extracting the positions where each component is connected as connection points in step (2) refers to identifying by calculating the distance between the surfaces of the components, and the position where the distance is 0 is identified as the connection point.

[0045] Furthermore, the step of converting the three-dimensional data of the skeleton structure into topological data in step (2) refers to taking the components in the skeleton structure as points in the topological network and the connection points as the edges of the connection points.

[0046] Furthermore, the step of extracting and learning the rules of the structural skeleton topological network in step (3) includes the connection relationship between Hana components and the number of connection points, the connection relationship between Hana components and Uni components, the connection relationship between Uni components and Taonao components and the number of connection points, and the connection relationship between Taonao components and the number of connection points.

[0047] Furthermore, the basic structure data of the yurt in step (4) includes the bottom diameter of the yurt, the height of the Hana wall, the diameter of the Taonao, the height of the bottom edge of the Taonao, and the total height of the yurt.

[0048] Furthermore, the yurt basic model database in step (4) is a database of yurt skeleton structure models storing the basic structure data of various yurts.

[0049] Furthermore, the structural mechanics analysis in step (5) refers to analyzing the stress, strain, and deformation of the structural skeleton under static loads through engineering simulation software such as ANSYS.

[0050] Furthermore, it is characterized in that the list and drawings of the prefabricated components in step (6) include the data of the type, quantity, diameter, and length of the prefabricated components.

Claims

1. A new intelligent generation method of yurt skeleton structure based on deep learning, characterized in that: The method comprises the following steps: Step (1), data collection of the 3D model of the yurt skeleton structure, Use laser scanning equipment to measure the skeleton of the yurt, collect point cloud data of the yurt skeleton structure, classify the point cloud data through a point cloud classification algorithm, obtain point cloud data of the Khana component, the Taonao component, and the Uni component, input the point cloud data into a 3D modeling software, generate a 3D model of the yurt skeleton structure consisting of the Khana, the Taonao, and the Uni, and build a 3D model library of the yurt skeleton structure; Step (2), topological transformation of the yurt skeleton structure and prefabricated components, Calculate the three-dimensional morphological properties of the Hana component, the Nao component, and the Uni component in step (1), including length, width, height, and cross-sectional shape, and construct a component three-dimensional morphological database; identify and extract the positions where each component is connected as the connection points, and according to the connection relationship between the components and the connection points, convert the three-dimensional data of the skeleton structure into topological data, and connect the connection points with the corresponding component data in the component three-dimensional morphological database to establish a sample topological data set, of which the training data set accounts for 80% and the test data set accounts for 20%; Step (3), constructing a skeleton structure generation model based on graph neural network, Input the training data set in step (2) into a deep learning workstation equipped with a graph neural network deep learning model, extract and learn the rules of the structural skeleton topology network, and construct a yurt skeleton structure generation model; Step (4), generation of the yurt skeleton structure model and matching generation of prefabricated components, According to the basic structural data of the yurt, a yurt basic model database is established, a yurt skeleton structure generation system is constructed, and the area and height data of the yurt to be built are input into the system. The yurt skeleton structure generation system can automatically match the basic model with the smallest error with the input area and height data, and input the basic model into the deep learning workstation in step (3), automatically generate a skeleton structure model, and provide data information and required quantity information of the three-dimensional model of standardized prefabricated components constituting the Khana component, the Nao component, and the Uni component; Step (5), stability test of the yurt skeleton structure model, Input the skeleton structure model generated in step (4) into the engineering simulation software to perform structural mechanics analysis. For the skeleton structure that generates structural displacement and deformation, return to step (4) to regenerate the skeleton structure model; Step (6), component drawing output, Export the required prefabricated component list and drawings into Excel and CAD format files for the next step of procurement and construction.

2. According to the novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1, it is characterized in that: The point cloud classification algorithm in step (1) is a deep learning algorithm trained by a point cloud data set, which can automatically classify unprocessed point cloud data.

3. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The step (2) of identifying and extracting the positions where the components are connected as the connection points refers to identifying by calculating the distance between the surfaces of the components, and identifying the position where the distance is 0 as the connection point.

4. According to the novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1, it is characterized in that: The step (2) of converting the three-dimensional data of the skeleton structure into topological data means that the components in the skeleton structure are points in the topological network and the connecting points are edges of the connecting points.

5. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The extraction and learning of the rules of the structural skeleton topological network in step (3) include the connection relationship and the number of connection points between Hana components, the connection relationship between Hana components and Uni components, the connection relationship and the number of connection points between Uni components and brain-mounted components, and the connection relationship and the number of connection points between brain-mounted components.

6. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The basic structural data of the yurt in step (4) include the bottom diameter of the yurt, the height of the Khana wall, the diameter of the yurt, the height of the bottom edge of the yurt and the total height of the yurt.

7. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The yurt basic model database in step (4) is a database of yurt skeleton structure models storing basic structural data of various yurts.

8. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The structural mechanics analysis in step (5) refers to analyzing the stress, strain and deformation of the structural skeleton under static load through engineering simulation software.

9. The novel intelligent generation method of yurt skeleton structure based on deep learning according to claim 1 is characterized in that: The prefabricated component list and drawings in step (6) include data on the type, quantity, diameter and length of the prefabricated components.

Images