An intelligent adaptation method for local updating of Monte building components

Abstract

The present invention discloses an intelligent adaptation method suitable for partial updating of Monte building components, and relates to the technical field of building component updating. The key points of the technical solution are: comprising the following steps: S1: construction of a target Monte building atlas; S2: construction of a Monte building case atlas; S3: case matching of components to be updated in the target building; S4: interactive display of updated buildings; S5: on-site construction correction. The present invention can not only improve the efficiency and accuracy of updating, but also improve the performance of buildings, extend their lifespan, and make buildings more sustainable and competitive. In this way, through intelligent matching and updating, the city's historical and cultural heritage can be better protected, and the sustainable development and improvement of urban buildings can be promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of building component renewal and adaptation, and more specifically, to an intelligent adaptation method suitable for local renewal of Monte building components. Background Art

[0002] In the process of urban renewal, the partial renewal of characteristic building components is not only to maintain the overall appearance of the city and continue the city's historical culture, but also to adapt to urban development and improve the functionality and economy of buildings. Through partial renewal, the damaged parts of old buildings can be repaired, their performance can be improved, their lifespan can be extended, and while retaining the original characteristics of the building, modern design concepts and technologies can be introduced to make the building more in line with contemporary needs.

[0003] Although the traditional partial building renovation method relies on the experience and professional knowledge of designers, it often has limitations and is difficult to fully consider the impact of various factors on building components. In addition, it lacks sufficient data support. Especially for Monte buildings, which have diverse elements and complex combinations, the traditional method is even more difficult to achieve a refined renovation plan.

[0004] Therefore, with the help of intelligent matching and updating methods and technologies, various factors can be considered more comprehensively and scientifically to achieve more accurate and personalized updates of building components; intelligent matching and updating methods can systematically analyze the structure, function, materials and other factors of building components through data analysis, simulation calculations and other means, so as to design more scientific and reasonable update plans. This method can not only improve the efficiency and accuracy of updates, but also improve the performance of buildings, extend their lifespan, and make buildings more sustainable and competitive. In this way, through intelligent matching and updating, the city’s historical and cultural heritage can be better protected, and the sustainable development and improvement of urban buildings can be promoted. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent adaptation method suitable for the partial renewal of Monte building components, so as to solve the problems in the prior art of the renewal of traditional characteristic buildings, such as the lack of matching case references, the strong subjectivity of the renewal and transformation process, the lack of scientific rationality and data quantification support, and the inability to verify the rationality of the updated building.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions: an intelligent adaptation method suitable for local updating of Monte building components, comprising the following steps:

[0007] S1: Obtain the basic attributes, image attributes and environmental attributes of each entity and its relationship of the Monte target building, and construct the Monte target building atlas;

[0008] S2: Obtain the models, images and text data of the implemented Monte architectural cases, convert them into basic attributes, image attributes and environmental attributes of each entity and its relationship, and construct a Monte architectural case map;

[0009] S3: searching for a subgraph to be updated in the Monte target building atlas described in step S1, identifying the same subgraph structure in the Monte building case atlas described in step S2 as the subgraph to be updated, obtaining a candidate subgraph, screening the candidate subgraphs having the same environmental attributes as the subgraph to be updated, and calculating the comprehensive similarity between the subgraph to be updated and the candidate subgraphs;

[0010] S4: displaying the entity models of the candidate sub-graphs in descending order of comprehensive similarity as described in step S3, and determining the updated model of the Monte target building to obtain an updated work manual;

[0011] S5: construct the updated model described in step S4 on site, record its wind load resistance value and thermal conductivity, and compare it with the local building construction standards. If they are not met, repeat step S4 to re-determine the updated model until the updated building fully meets the standards.

[0012] By adopting the above technical scheme, the present invention constructs building atlases and case atlases to quantify and compare the basic attributes, image attributes and environmental attributes of the building, thereby reducing subjectivity and increasing the scientificity and accuracy of the updating process; the present invention also utilizes the method of atlas construction and comprehensive similarity calculation to comprehensively consider various attribute information of the building, thereby improving the matching degree and selection accuracy of the updating model, and ensuring that the updated building is perfectly matched with the original building; finally, the present invention ensures the safety and sustainability of the updated building and guarantees the quality of the building through on-site construction and testing of physical performance indicators such as wind load resistance and thermal conductivity, as well as comparison and testing with local building construction standards; the present method solves the problems of lack of matching case references, strong subjectivity and rationality verification in the updating of traditional characteristic buildings through scientific and reliable guidance, has high practical value and application prospects, and can provide effective support and guidance for the local updating of Monte building components.

[0013] The present invention is further configured as follows: the specific form of constructing the graph in step S1 and step S2 is to construct the graph in the form of triples, namely, entity 1-relationship-entity 2.

[0014] By adopting the above technical scheme, the present invention constructs a map in the form of triples, which can clearly express the relationship between building attributes, making the information organization structure clear, easy to understand and manage. The map in the form of triples can normalize and standardize the data, facilitate subsequent data processing and analysis, and improve the quality and reliability of the data. The map in the form of triples can easily expand and update the data, so that the system has high flexibility and scalability.

[0015] The present invention is further configured as follows: the specific method for obtaining the candidate subgraph in step S3 is: obtaining the candidate subgraph by identifying the subgraph structure identical to the Monte Carlo architectural case map described in step S2 through a subgraph isomorphism matching algorithm, wherein the subgraph isomorphism matching algorithm specifically determines whether two subgraphs are isomorphic by recursively searching for node matching of the two subgraphs.

[0016] By adopting the above technical solution, the present invention first needs to determine the nodes with the same structure in the two subgraphs, and recursively search the two subgraphs to match each node to determine whether they are isomorphic; in addition to node matching, it is also necessary to consider the connection relationship between nodes, that is, edge matching. By comparing the connection of the edges in the two subgraphs, it is further determined whether they are isomorphic, and based on the matching of nodes and edges, it is determined whether the two subgraphs are isomorphic, that is, whether they have the same structure and connection method. If the isomorphism condition is met, it is considered that the same subgraph structure has been found;

[0017] The present invention can accurately identify the same structure in two subgraphs by recursively searching for the matching of nodes and edges, thereby improving the accuracy of matching. The algorithm can efficiently match two subgraphs and quickly determine whether they are isomorphic by comparing nodes and edges, thus saving time and resources. The subgraph isomorphic matching algorithm is suitable for various types of graph structure matching and has strong versatility and applicability.

[0018] The present invention is further configured as follows: the method for calculating the comprehensive similarity in step S3 refers to adding the basic attribute similarity and the image attribute similarity, that is, the comprehensive similarity S=basic attribute similarity SA+image attribute similarity SP.

[0019] By adopting the above technical solution, the present invention selects nodes one by one from the node sets of the two subgraphs, and determines whether they match by comparing the attribute information of the nodes. If the attributes of the two nodes are consistent, the match is considered successful; on the basis of determining the node matching relationship, the edge matching is considered. By comparing the connection relationship and attribute information of the edges, it is determined whether there is consistency of the edges. Based on the matching results of the nodes and edges, the subgraph matching is recursively expanded to search until the isomorphic relationship of the entire subgraph structure is determined; the present invention can achieve accurate identification and matching of the subgraph structure through one-to-one matching of nodes and edges, ensuring that the obtained candidate subgraphs are consistent with the case map. The algorithm adopts a recursive search method, which can efficiently find the same subgraph structure in the candidate subgraphs, reduce matching time and computational complexity, and the algorithm can be expanded and improved according to actual needs. It is suitable for different types of subgraph matching problems and has good scalability and adaptability.

[0020] The present invention is further configured as follows: the basic attribute similarity SA is obtained by calculating the angle cosine similarity between the subgraph to be updated and the candidate subgraph in clustering, specifically:

[0021] For any two vectors x1=(x 11 ,x 12 …x 1n ),x2=x 21 ,x 22 …x 2n ), the cosine of the angle between the two vectors S i for:

[0022]

[0023] The basic attribute similarity SA is:

[0024]

[0025] By adopting the above technical solution, the present invention converts the basic attribute features of the subgraph to be updated and the candidate subgraph into vector form, each attribute feature corresponds to a component in the vector, and the similarity between them is expressed by calculating the cosine value of the angle between the two vectors. The angle cosine similarity value is between -1 and 1, and the larger the value, the higher the similarity. The angle cosine similarity can be used in cluster analysis to calculate the similarity between the subgraph to be updated and the candidate subgraph, so as to determine the degree of similarity between them; the angle cosine similarity calculation method of the present invention is simple and intuitive, can accurately measure the similarity between two vectors, and is convenient for calculation and understanding. The angle cosine similarity is a similarity measurement method with clear physical meaning in mathematics, which can clearly express the degree of similarity between the subgraph to be updated and the candidate subgraph. The angle cosine similarity can be applied to various types of vector form data, and has strong versatility and applicability.

[0026] The present invention is further configured as follows: the specific method for calculating the image attribute similarity SP is: first, the image is scaled to 8*8 pixels and grayed, each pixel is binarized according to the average value of the image, a hash value is calculated, a value greater than the average value is set to 1, and a value less than the average value is set to 0, a hash value is generated, and based on the two hash values, the similarity SP of the two images is obtained.

[0027] By adopting the above technical scheme, the present invention can quickly calculate the hash value of the image by simply processing and binarizing the image, and perform similarity comparison. The calculation speed is fast, and the calculation method of the image hash value is not affected by transformations such as rotation and scaling. The image processing has strong robustness, and through the comparison of hash values, the similarity between two images can be intuitively shown, which is convenient for image similarity comparison and matching.

[0028] The present invention is further configured as follows: Step S5 records the wind load resistance value and thermal conductivity coefficient through wind resistance performance testing and thermal insulation performance testing.

[0029] The present invention is further configured as follows: the wind resistance performance test refers to setting a test wind load value, simulating the actual wind action form in the laboratory, and cyclically loading the test piece. If the building is damaged or functionally destroyed, the test is stopped accordingly and the set wind load value is recorded.

[0030] By adopting the above technical scheme, the present invention can accurately evaluate the bearing capacity of building structures under wind loads by simulating the actual wind load action form in the laboratory, providing an important reference for building design and improvement. Through wind resistance testing, potential problems and weaknesses in the building structure can be discovered in time to ensure the safety of the building under extreme weather conditions such as wind disasters. By recording the wind load value and building performance during the test, objective and reliable data support can be provided, providing a basis for the improvement of the building structure and safety assessment.

[0031] In summary, the present invention has the following beneficial effects:

[0032] 1. The present invention constructs a target Monte building atlas by finely modeling the structural elements of the target building to be updated in layers, which can accurately determine and identify the components to be updated of the building; further, the model data of the Monte building case is split and calculated to obtain basic attributes, and the text data is extracted to obtain environmental attributes, and the Monte building case atlas is constructed in combination with the image attributes, which is convenient for query and improves the accuracy of subsequent case identification and matching;

[0033] 2. The present invention compares the similarity of comprehensive multi-dimensional attributes of the cases matched to the components to be updated by screening the same environmental attributes, calculating the similarity of basic attributes, and calculating the similarity of image attributes, and finds the case with the highest similarity to the attributes of the components to be updated, thereby maximizing the accuracy of case matching and optimizing the calculation process, thereby improving the efficiency of the entire case matching process;

[0034] 3. The present invention improves the operability of the output results by outputting an update work manual for the updated building, which not only provides guidance for the on-site construction work, but also provides a basis for the subsequent maintenance and repair of the building. At the same time, the updated building is constructed on-site, and its wind load resistance value and thermal conductivity coefficient are compared with the local construction standards through wind resistance performance testing and thermal insulation performance testing, which enhances the integration of the building and the environment and improves its local acceptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a flow chart of the method of the present invention; DETAILED DESCRIPTION

[0036] The following is combined with Figure 1 The present invention is described in further detail.

[0037] Embodiment: An intelligent adaptation method for local updating of Monte building components, such as Figure 1 As shown, the following steps are included:

[0038] S1: Target Monte Architectural Atlas Construction

[0039] Select the target building, use a laser 3D scanner with a single-frame measurement accuracy of ≤±0.01mm to perform 3D scanning on the target building, and model 15 components of three categories of Monte architecture, namely, wooden frame system (taonao, uni, hana, cowhide nails, wooden doors, pillars), felt system (midge felt, top felt, surrounding felt, top decoration, foot felt), and rope system (surrounding rope, falling rope, pressure rope, binding rope) in layers. Take the 15 components as entities of the knowledge graph, and the connection mode between the components as relationships. Calculate the basic attributes of each entity and relationship. The basic attributes refer to recording the characteristic values ​​that can be described by each entity (as shown in Table 1 below) and the characteristic values ​​that can be described by each association relationship (as shown in Table 2 below);

[0040] Table 1 Description of basic attribute indicators of entities

[0041]

[0042]

[0043] Table 2 Description of basic attribute indicators of relationships

[0044]

[0045]

[0046] Obtain image data of the target Monte building and the buildings within 500m of its surroundings as the image attributes of the entity, and obtain the annual average wind speed, annual average precipitation, and annual average sunshine duration of the location of the target Monte building as the environmental attributes of the entity; construct a target Monte building map in the form of a triple (entity 1-relationship-entity 2) for each entity, relationship, basic attribute, image attribute, and environmental attribute.

[0047] S2: Construction of Monte Architectural Case Map

[0048] Obtain models, images and text data of 500 implemented Monte architectural cases from the planning administrative department, split the model data into 15 entities, split the connection mode into relationships, and calculate the basic attributes of the entities and relationships respectively; use the image data as image attributes, use the text data to obtain environmental attributes using the text information extraction model, and call step S2 to build a Monte architectural case map;

[0049] The above-mentioned text information extraction model refers to the use of the text information extraction model to obtain triples of environmental attributes; wherein, the text information extraction model includes a named entity recognition model based on deep learning and a supervised relationship extraction model. The named entity recognition model uses the BIO annotation rules to tokenize the segmented text and identify entities with specific meanings in the text. On this basis, 10% of the sentences in the text are selected to manually annotate the relationship between entities, and a binary classifier is obtained using supervised learning to determine whether there is a specific relationship between two unlabeled entities. This is used as a relationship extraction model to obtain the relationship in 90% of the unlabeled text.

[0050] S3: Case matching of target building components to be updated

[0051] In the target Monte Carlo architectural graph, search for the entities and three types of attributes to be updated and the entities and three types of attributes directly associated with them, output the subgraph to be updated, where direct association means that there is only one relationship between two entities, convert the subgraph to be updated into a representation vector, use the subgraph isomorphism matching algorithm to identify the same subgraph structure in the case graph to obtain the candidate subgraph, screen the candidate subgraphs with the same environmental attributes as the subgraph to be updated, use the clustering algorithm to obtain the similarity SA of the basic attributes of the subgraph to be updated and the candidate subgraph, use the mean hashing algorithm to obtain the similarity SP of the image attributes, and add the two together as the comprehensive similarity S;

[0052] The above-mentioned use of the subgraph isomorphism matching algorithm to identify the same subgraph structure in the case graph to obtain the candidate subgraph specifically refers to recursively searching for node matches of two subgraphs to determine whether they are isomorphic, which specifically includes two parts: forward search and reverse search. In the forward search, it is checked whether the adjacent nodes match; in the reverse search, it backtracks to find other possible matching methods until all matching cannot be performed anymore;

[0053] The specific calculation method of the above comprehensive similarity S is: add the basic attribute similarity and the image attribute similarity, that is, the comprehensive similarity S = basic attribute similarity SA + image attribute similarity SP;

[0054] The basic attribute similarity SA is obtained by calculating the angle cosine similarity between the subgraph to be updated and the candidate subgraph in the cluster, specifically:

[0055] For any two vectors x1=(x 11 ,x 12 …x 1n ),x2=x 21 ,x 22 …x 2n ), the cosine of the angle between the two vectors S i for:

[0056]

[0057] The basic attribute similarity SA is:

[0058]

[0059] Image attribute similarity SP uses the mean hash algorithm. First, the image is scaled to 8*8 pixels and grayed. Each pixel is binarized according to the average value of the image, and the hash value is calculated. The value greater than the average value is set to 1, and the value less than the average value is set to 0. The similarity SP of the two images is calculated based on the two hash values.

[0060] S4: Updated building interactive display

[0061] The matching candidate sub-images are output in descending order of comprehensive similarity, and the entity models of the candidate sub-images are displayed on the holographic sandbox. The updated model of the target building is determined through the user's interactive selection, and the updated work manual is output in combination with the style, size and other information of the updated model.

[0062] S5: On-site construction and correction

[0063] The updated building is constructed on site, and its wind load resistance value and thermal conductivity are recorded through wind resistance performance test and thermal insulation performance test, and compared with the local building construction standards. If it is not met, a candidate subgraph is reselected and step S4 is repeated until the updated building fully meets the standards;

[0064] The wind resistance test refers to setting the wind load value, simulating the actual wind action in the laboratory, and cyclically loading the test specimens. If the building is damaged or functionally damaged, the test is stopped and the set wind load value is recorded. The thermal insulation test refers to applying a certain heat flow and pressure to the building envelope structure, testing the thickness of the sample and the temperature difference between the heat plates, and recording the thermal conductivity.

[0065] Working principle: The present invention first constructs the Monte target building map and the Monte building case map through steps S1 and S2, and simultaneously obtains the basic attributes, image attributes and environmental attributes of each entity and the relationship between them; next, through step S3, the subgraph to be updated is searched in the Monte target building map, and the same subgraph structure in the Monte building case map is identified through the subgraph isomorphism matching algorithm to obtain the candidate subgraph; then, a comprehensive similarity calculation is performed based on the environmental attributes of the candidate subgraph and the subgraph to be updated, and the candidate subgraph with similar environmental attributes to the subgraph to be updated is screened out;

[0066] In step S4, the entity model is displayed in order of the comprehensive similarity of the candidate subgraphs, and the updated model of the target building in Monte is determined, and an update work manual is generated; finally, in step S5, the updated model is built on site, and the wind load resistance and thermal conductivity of the updated building are recorded and compared with the local building construction standards. If the updated building does not meet the standards, it is necessary to re-determine the updated model in step S4, and repeatedly optimize until the updated building fully meets the standards;

[0067] The present invention determines the best update model by establishing a similarity comparison between building graphs and sub-graph structures, combining the screening of environmental attributes, and ensures that the updated buildings meet local building standards through comprehensive similarity calculation and on-site construction. At the same time, the efficiency and accuracy of design and construction are improved. By gradually optimizing the update model, this method can effectively realize the intelligent adaptation of local updates of Monte building components.

[0068] This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

Claims

1. An intelligent adaptation method suitable for local updating of Monte-Tec components, characterized by: The following steps are involved: S1: Obtain the basic attributes, image attributes and environmental attributes of each entity and its relationship of the Monte target building, and construct the Monte target building atlas; S2: Obtain the models, images and text data of the implemented Monte architectural cases, convert them into basic attributes, image attributes and environmental attributes of each entity and its relationship, and construct a Monte architectural case map; S3: searching for a subgraph to be updated in the Monte target building atlas described in step S1, identifying the same subgraph structure in the Monte building case atlas described in step S2 as the subgraph to be updated, obtaining a candidate subgraph, screening the candidate subgraphs having the same environmental attributes as the subgraph to be updated, and calculating the comprehensive similarity between the subgraph to be updated and the candidate subgraphs; The method of calculating the comprehensive similarity in step S3 is to add the basic attribute similarity and the image attribute similarity, that is, the comprehensive similarity S = basic attribute similarity SA + image attribute similarity SP; S4: displaying the entity models of the candidate sub-graphs in descending order of comprehensive similarity as described in step S3, and determining the updated model of the Monte target building to obtain an updated work manual; S5: construct the updated model described in step S4 on site, record its wind load resistance value and thermal conductivity, and compare it with the local building construction standards. If they are not met, repeat step S4 to re-determine the updated model until the updated building fully meets the standards.

2. The intelligent adaptation method for local updating of Monte building components according to claim 1 is characterized in that: The specific form of constructing the graph in step S1 and step S2 is to construct the graph in the form of triples, namely entity 1-relationship-entity 2.

3. The intelligent adaptation method for partial updating of Monte building components according to claim 1 is characterized in that: The specific method for obtaining the candidate subgraph in step S3 is: obtaining the candidate subgraph by identifying the subgraph structure identical to the Monte Architectural Case Graph described in step S2 through a subgraph isomorphism matching algorithm, wherein the subgraph isomorphism matching algorithm specifically determines whether two subgraphs are isomorphic by recursively searching for node matching of the two subgraphs.

4. The intelligent adaptation method for partial updating of Monte building components according to claim 1 is characterized in that: The basic attribute similarity SA is obtained by calculating the angle cosine similarity between the subgraph to be updated and the candidate subgraph in the cluster, specifically: For any two vectors x1=(x 11 ,x 12 …x 1n ),x2=x 21 ,x 22 …x 2n ), the cosine of the angle between the two vectors S i for: The basic attribute similarity SA is:

5. The intelligent adaptation method for local updating of Monte building components according to claim 4 is characterized in that: The specific method of calculating the image attribute similarity SP is as follows: first, scale the image to 8*8 pixels and grayscale it, binarize each pixel according to the average value of the image, calculate the hash value, set the value greater than the average value to 1, and set the value less than the average value to 0, generate a hash value, and obtain the similarity SP of the two images based on the two hash values.

6. The intelligent adaptation method for partial updating of Monte building components according to claim 1 is characterized by: In step S5, the wind load resistance value and thermal conductivity coefficient are recorded through wind resistance performance testing and thermal insulation performance testing.

7. The intelligent adaptation method for partial updating of Monte building components according to claim 6 is characterized by: The wind resistance test refers to setting the test wind load value, simulating the actual wind action form in the laboratory, and cyclically loading the test piece. If the building is damaged or functionally destroyed, the test is stopped accordingly and the set wind load value is recorded.

Images