A digital twin model construction and processing method and system for special-shaped buildings

Abstract

The present invention relates to the field of architectural modeling technology, and in particular to a method and system for constructing and processing a digital twin model of a special-shaped building. The method comprises obtaining architectural geometric information of a target special-shaped building, determining the first type of architectural components, the second type of architectural components, and the third type of architectural components of the target special-shaped building based on the architectural geometric information, setting a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of architectural components, the second type of architectural components, and the third type of architectural components, performing a differentiated point cloud scan on the target special-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density, and obtaining a point cloud data set; and constructing a digital twin model of the target special-shaped building based on the point cloud data set to obtain a target digital twin model. While ensuring the modeling accuracy of complex morphological components, the method achieves the technical effect of reducing the overall data volume and processing cost.

Description

Technical Field

[0001] The present invention relates to the technical field of building modeling, and in particular to a method and system for constructing and processing a digital twin model of a special-shaped building. Background Art

[0002] With the development of digitalization in the construction industry, special-shaped buildings such as curved structures and non-standard buildings are increasing as people's demand for architectural aesthetics and functionality grows. This poses new challenges to architectural modeling technology. The commonly used technology for architectural modeling is digital twin technology. As a core tool for the management of the entire life cycle of buildings, digital twin technology needs to convert physical buildings into virtual models through means such as point cloud scanning. Traditional architectural modeling methods usually use a uniform scanning density to scan the entire building, and then build a model based on the full amount of point cloud data. Although it can ensure the accuracy of complex components, it will generate a large amount of redundant data, increase storage and processing costs and modeling time; if low-density scanning is used, the key details of special-shaped components may be lost, resulting in model distortion. Existing architectural modeling technology has the technical problem that the scanning strategy cannot match the actual modeling requirements of each component, and it is difficult to take into account the accuracy and economy of the overall model. Summary of the Invention

[0003] The present invention aims to solve the technical problem in the existing technology that it is difficult to strike a balance between the accuracy and economy of the overall model, and provides a method and system for constructing and processing a digital twin model of an irregular-shaped building.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] In a first aspect, the present invention provides a method for constructing and processing a digital twin model of an irregularly shaped building, comprising: obtaining architectural geometric information of a target irregularly shaped building, and determining a first type of building components, a second type of building components, and a third type of building components of the target irregularly shaped building based on the architectural geometric information; setting a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density; performing a differentiated point cloud scan on the target irregularly shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud data set; and constructing a digital twin model of the target irregularly shaped building based on the point cloud data set to obtain a target digital twin model.

[0006] Optionally, determining the first category, second category, and third category of building components of the target special-shaped building based on the building geometric information includes: determining a plurality of building components based on the building geometric information, and extracting morphological characteristic parameters of each of the building components; determining a component abnormality of each of the building components based on the morphological characteristic parameters of each of the building components; comparing the component abnormality of each of the building components with a preset first abnormality threshold and a second abnormality threshold, wherein the second abnormality threshold is greater than the first abnormality threshold; when the component abnormality is less than the first abnormality threshold, classifying the corresponding building component as a first category of building component; when the component abnormality is greater than or equal to the first abnormality threshold and less than the second abnormality threshold, classifying the corresponding building component as a second category of building component; and when the component abnormality is greater than or equal to the second abnormality threshold, classifying the corresponding building component as a third category of building component.

[0007] The method of determining the component abnormality of each building component based on the morphological characteristic parameters of each building component includes: extracting a first building component from the multiple building components and obtaining the morphological characteristic parameters of the first building component as the first morphological characteristic parameters; calling an abnormality identifier and processing the first morphological characteristic parameters by the abnormality identifier to obtain the first component abnormality of the first building component; and obtaining the component abnormality of the remaining building components in the same manner as that of obtaining the first component abnormality of the first building component to obtain the component abnormality of each building component.

[0008] Among them, the steps of constructing the abnormality identifier include: collecting multiple sample building components, acquiring morphological feature parameters of the multiple sample building components, and obtaining a sample morphological feature parameter set; marking the component abnormality of the multiple sample building components according to the sample morphological feature parameter set to obtain a sample component abnormality set; and using machine learning to train and obtain the abnormality identifier based on the sample morphological feature parameter set and the sample component abnormality set.

[0009] Among them, after obtaining the first component abnormality of the first building component, the method also includes: calling a historical component library, the historical component library including multiple historical building components and morphological characteristic parameters and occurrence frequencies of each of the historical building components; searching and matching the morphological characteristic parameter pairs of each of the historical building components in the historical component library based on the first morphological characteristic parameters to determine multiple first matching components; counting the occurrence frequencies of the multiple first matching components to obtain a first total occurrence frequency; determining a first abnormality correction coefficient based on the first total occurrence frequency and a preset basic occurrence frequency; correcting the first component abnormality by using the first abnormality correction coefficient, and using the correction result as the first component abnormality.

[0010] The method further comprises: receiving a point cloud scanning density of a standard component; setting the point cloud scanning density of the standard component as the first point cloud scanning density of the first category of building components; obtaining a second category component abnormality concentration value based on the component abnormality of each building component in the second category, and obtaining the second point cloud scanning density by combining the point cloud scanning density of the standard component; obtaining a third category component abnormality concentration value based on the component abnormality of each building component in the third category, and obtaining the third point cloud scanning density by combining the point cloud scanning density of the standard component.

[0011] Optionally, performing differentiated point cloud scanning on the target special-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud dataset includes: determining spatial position information of the first type of building components, the second type of building components, and the third type of building components in the target special-shaped building; dividing the target special-shaped building into a first scanning area, a second scanning area, and a third scanning area according to the spatial position information; performing point cloud data collection in the first scanning area according to the first point cloud scanning density to obtain a first point cloud dataset; performing point cloud data collection in the second scanning area according to the second point cloud scanning density to obtain a second point cloud dataset; performing point cloud data collection in the third scanning area according to the third point cloud scanning density to obtain a third point cloud dataset; and aggregating the first point cloud dataset, the second point cloud dataset, and the third point cloud dataset to obtain the point cloud dataset.

[0012] In a second aspect, the present invention provides a digital twin model construction and processing system for special-shaped buildings, comprising:

[0013] a geometric information acquisition module, configured to acquire architectural geometric information of a target special-shaped building, and determine first-category architectural components, second-category architectural components, and third-category architectural components of the target special-shaped building based on the architectural geometric information;

[0014] a point cloud density allocation module, configured to set a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density;

[0015] a point cloud scanning execution module, configured to perform a differentiated point cloud scan on the target irregular-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud data set;

[0016] The digital model construction module is used to construct a digital twin model of the target special-shaped building according to the point cloud data set to obtain a target digital twin model.

[0017] By implementing the present invention, it is possible to obtain architectural geometric information of a target special-shaped building, determine the first-category architectural components, the second-category architectural components, and the third-category architectural components of the target special-shaped building based on the architectural geometric information, realize differentiated identification of architectural components, clarify the morphological complexity of different components, provide a classification basis for subsequent differentiated setting of scanning density, and avoid waste of resources or insufficient precision caused by adopting a unified processing method for all components.

[0018] By implementing the present invention, it is possible to set a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density based on the first, second, and third types of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density. A higher scanning density is assigned to components with complex shapes and high abnormalities to ensure detail capture; the density is lowered for simple components to reduce data redundancy, balance model accuracy and data volume, and optimize scanning efficiency.

[0019] By implementing the present invention, it is possible to perform differentiated point cloud scanning on the target irregular-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density, obtain a point cloud data set, and ensure high-precision data collection in high-abnormality areas through spatial partition scanning, while reducing the data collection cost in low-abnormality areas, thereby improving the pertinence and efficiency of the overall scanning;

[0020] By implementing the present invention, it is possible to construct a digital twin model of the target special-shaped building based on the point cloud data set, and obtain a target digital twin model. Based on differentiated point cloud data, while ensuring the modeling accuracy of complex components, model redundancy caused by overall high-density scanning is avoided, and finally an accurate and efficient digital twin model is obtained to meet the special needs of special-shaped buildings for detail restoration.

[0021] In summary, by implementing the present invention, it is possible to achieve the technical effect of reducing the overall data volume and processing costs while ensuring the modeling accuracy of complex morphological components. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1A schematic diagram of a process for constructing a digital twin model of a special-shaped building provided by the present invention;

[0023] Figure 2 This is a structural schematic diagram of a digital twin model construction and processing system for special-shaped buildings provided by the present invention.

[0024] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0025] Geometric information acquisition module 11, point cloud density allocation module 12, point cloud scanning execution module 13, digital model construction module 14. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Therefore, features specified as "first" or "second" may explicitly or implicitly include one or more features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0028] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0029] Example 1, as Figure 1 As shown, an embodiment of the present invention provides a method for constructing a digital twin model of a special-shaped building, including:

[0030] S100: Acquire architectural geometric information of a target special-shaped building, and determine first-category architectural components, second-category architectural components, and third-category architectural components of the target special-shaped building based on the architectural geometric information;

[0031] S200: Setting a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density;

[0032] S300: performing a differentiated point cloud scan on the target irregular-shaped building based on the first point cloud scan density, the second point cloud scan density, and the third point cloud scan density to obtain a point cloud dataset;

[0033] S400: Constructing a digital twin model of the target special-shaped building based on the point cloud dataset to obtain a target digital twin model.

[0034] In step S100 of the embodiment of the present application, determining the first type of building components, the second type of building components, and the third type of building components of the target special-shaped building based on the building geometric information includes:

[0035] determining a plurality of building components according to the building geometric information, and extracting morphological characteristic parameters of each of the building components;

[0036] determining a component abnormality degree of each of the building components based on the morphological characteristic parameters of each of the building components;

[0037] Comparing the component abnormality of each of the building components with a preset first abnormality threshold and a second abnormality threshold, wherein the second abnormality threshold is greater than the first abnormality threshold;

[0038] When the component abnormality is less than the first abnormality threshold, classifying the corresponding building component as a first type of building component;

[0039] When the abnormality of a component is greater than or equal to the first abnormality threshold and less than the second abnormality threshold, classifying the corresponding building component as a second type of building component;

[0040] When the component abnormality is greater than or equal to the second abnormality threshold, the corresponding building component is classified as a third type of building component.

[0041] In the embodiment of the present application, the core purpose of this step is to differentiate the various components of the target special-shaped building and divide them into three categories according to the complexity of the components' shapes, so as to provide an accurate classification basis for subsequent differentiated point cloud scanning, thereby ensuring the modeling accuracy of complex components while avoiding the waste of resources caused by excessive scanning of simple components.

[0042] In a specific implementation process, it is first necessary to extract the morphological characteristic parameters of each of the building components to determine the component abnormality of each of the building components.

[0043] In step S100 of the embodiment of the present application, determining the abnormality of each building component based on the morphological characteristic parameters of each building component includes:

[0044] extracting a first building component from the plurality of building components, and obtaining a morphological characteristic parameter of the first building component as a first morphological characteristic parameter;

[0045] Retrieving an abnormality degree identifier, and processing the first morphological characteristic parameter by the abnormality degree identifier to obtain a first component abnormality degree of the first building component;

[0046] The component abnormality degrees of the remaining building components are obtained in the same manner as that of the first building component, thereby obtaining the component abnormality degree of each building component.

[0047] In this embodiment of the present application, extracting the morphological characteristic parameters of building components from the multiple building components is a fundamental step in determining the abnormality of a building component. Its core purpose is to provide specific input data for calculating the abnormality of a single building component. By extracting the morphological characteristic parameters of a specific building component, such as the first building component, as the raw basis for calculating the abnormality of that component by the abnormality identifier, this lays the foundation for quantifying the abnormality of all building components, ultimately supporting the classification of building components.

[0048] Specifically, one of the multiple building components of the target irregularly shaped building is selected as the processing target, namely the "first building component." Specific parameters reflecting the morphological characteristics of this selected first building component, such as curvature, dimensional deviation, non-standard angles, and surface complexity, are extracted from its corresponding architectural geometric information. These parameters are then integrated to form the "first morphological characteristic parameters."

[0049] For example, taking a "hyperbolic paraboloid roof component," a common feature in irregularly shaped buildings, the process for extracting its morphological characteristic parameters involves extracting curvature values ​​at different locations. For example, a hyperbolic paraboloid roof has a maximum curvature of 5.2 / m, a minimum curvature of 1.8 / m, and a standard deviation of 1.3. These parameters reflect the severity and complexity of the surface's curvature. Next, the actual dimensional deviation of the hyperbolic paraboloid roof is calculated by comparing it with the dimensional specifications of a standard flat roof. Examples include the overall height deviation from the standard flat roof, such as ±0.8m; and the offset of the edge contour from the standard rectangular outline, such as a maximum offset of 1.5m. These parameters reflect the dimensional differences between the component and the standard form. Next, non-standard angle values ​​are collected. Since the connection between the roof and the wall is not a standard 90° right angle, angle values ​​such as 82° and 97° are extracted to reflect the irregularities of the component's connection. Finally, the surface's irregular mesh proportion can be calculated by meshing. For example, among 100 sampling grids, the normal direction of 78 grids has an angle greater than 15° with the normal of the standard plane, which is used to quantify the overall morphological complexity of the surface.

[0050] After integrating the above parameters such as curvature, dimensional deviation, non-standard angle, and surface complexity, the "first morphological characteristic parameters" of the hyperbolic paraboloid roof component are formed, providing a specific quantitative basis for the subsequent calculation of its component abnormality.

[0051] Through this method, the morphological characteristic parameters of the target building components can be obtained.

[0052] After extracting the first morphological characteristic parameters of the first building component, it is necessary to call an abnormality degree identifier, and process the first morphological characteristic parameters by the abnormality degree identifier to obtain a first component abnormality degree of the first building component.

[0053] In step S100 of the embodiment of the present application, the step of constructing the abnormality identifier includes:

[0054] Collecting a plurality of sample building components, acquiring morphological characteristic parameters of the plurality of sample building components, and obtaining a sample morphological characteristic parameter set;

[0055] performing component abnormality identification on the plurality of sample building components according to the sample morphological characteristic parameter set to obtain a sample component abnormality degree set;

[0056] The abnormality identifier is trained and obtained based on the sample morphological feature parameter set and the sample component abnormality set using machine learning.

[0057] In the embodiment of the present application, the core purpose of constructing an anomaly identifier is to obtain a tool that can automatically and accurately calculate the abnormality of building components based on the morphological characteristic parameters of the building components, provide a reliable quantitative basis for the subsequent classification of building components, ensure that the classification standards are objective and consistent, and support the implementation of differentiated point cloud scanning strategies.

[0058] To obtain the abnormality identifier, it is first necessary to collect a plurality of sample building components, acquire morphological feature parameters of the plurality of sample building components, and obtain a set of sample morphological feature parameters as training samples.

[0059] The training samples, i.e., the source of the sample morphological feature parameter set, require the following: Various building components from different types of buildings should be selected as sample building components. These should include standard components such as rectangular beams and flat walls; slightly irregularly shaped components such as columns with small curvatures; and severely irregularly shaped components such as hyperbolic paraboloid roofs and twisted steel structures. This ensures that the samples cover a range of different abnormality levels.

[0060] Then, for each sample building component, its morphological characteristic parameters are extracted, including the aforementioned curvature, dimensional deviation, non-standard angle, surface complexity, etc., and these parameters are integrated to form a sample morphological characteristic parameter set.

[0061] Based on the sample morphological characteristic parameter set, the component abnormality of each sample building component is identified. For example, the closer the morphology is to the standard component, the lower the abnormality identification value; the further the morphology deviates from the standard component, the higher the abnormality identification value, ultimately forming a sample component abnormality set. Then, based on the sample morphological characteristic parameter set, the component abnormality of each sample building component is identified. For example, the closer the morphology is to the standard component, the lower the abnormality identification value; the further the morphology deviates from the standard component, the higher the abnormality identification value, ultimately forming a sample component abnormality set. The abnormality degree can be represented by a value between 0 and 1, with larger values ​​indicating higher abnormality.

[0062] A total of no less than 10,000 sample building component morphological characteristic parameters and abnormality are collected, and a training set and a validation set are divided into a ratio of 8:2 for training the abnormality identifier.

[0063] For the task type of the anomaly degree identifier, a multi-layer perceptron neural network can be used to build the anomaly degree identifier to handle the nonlinear relationship between complex morphological feature parameters and anomaly degree.

[0064] The anomaly detector consists of three layers: an input layer, a hidden layer, and an output layer. The number of neurons in the input layer matches the number of parameters in the sample morphological feature parameter set. For example, if only four parameters, such as curvature and dimensional deviation, are included, the input layer has four neurons. Two hidden layers are set, with the first layer containing 32 neurons and the second layer containing 16 neurons, for progressive feature extraction. The output layer consists of one neuron, and the output result is the component anomaly degree, ranging from 0 to 1, with 0 representing the closest to the standard component and 1 representing the highest degree of anomaly.

[0065] For the anomaly detector, use the ReLU function for the hidden layer and the Sigmoid function for the output layer. Adam is used as the optimizer, with a learning rate of 0.001. Mean squared error is used as the loss function. Each training step uses 32 input samples. The training samples used in the previous step are the same. A total of 100 training rounds are used.

[0066] When the loss function value of the validation set does not decrease for 10 consecutive rounds and the decrease is less than 0.0001, the training is stopped. At this time, the model reaches a convergence state and the anomaly identifier is obtained.

[0067] Finally, an abnormality degree identifier is retrieved, and the first morphological characteristic parameter is processed by the abnormality degree identifier to obtain a first component abnormality degree of the first building component.

[0068] For example, for a first building component, its structure is a hyperbolic paraboloid roof, and its first morphological characteristic parameters are a maximum curvature of 5.2 / m, a dimensional deviation of ±0.8m, non-standard connection angles of 82° and 97°, and an irregular grid accounting for 78% of the curved surface; after calling the anomaly identifier, the above parameters are input, and the identifier uses internal neural network calculations to output that the anomaly of the first component of the roof is 0.85.

[0069] Next, the component abnormality degrees of the remaining building components are obtained in the same manner as that of obtaining the first component abnormality degree of the first building component, thereby obtaining the component abnormality degree of each building component.

[0070] For example, a second building component is a column with a slight curvature. Its morphological characteristic parameters are: curvature 1.3 / m, dimensional deviation ±0.2m, standard angle ratio 90%, and irregular curved mesh ratio 15%. Inputting the above method into the anomaly detector yields a component anomaly index of 0.32.

[0071] For example, a third building component is a rectangular beam, a standard component. Its morphological characteristic parameters are: curvature 0 / m, dimensional deviation 0m, connection angle 90°, and irregular mesh ratio 0%. After inputting these parameters into the anomaly detector, the component anomaly score is 0.1.

[0072] Through the above method, the component abnormality degree of the building component can be obtained.

[0073] In step S100 of the embodiment of the present application, after obtaining the first component abnormality degree of the first building component, the method further includes:

[0074] Retrieving a historical component library, wherein the historical component library includes a plurality of historical building components and morphological characteristic parameters and occurrence frequencies of each of the historical building components;

[0075] Searching and matching the morphological characteristic parameter pairs of each of the historical building components in the historical component library based on the first morphological characteristic parameter to determine a plurality of first matching components;

[0076] Counting the occurrence frequencies of the plurality of first matching components to obtain a first total occurrence frequency;

[0077] determining a first abnormality correction coefficient according to the first total occurrence frequency and a preset basic occurrence frequency;

[0078] The first component abnormality is corrected using the first abnormality correction coefficient, and the correction result is used as the first component abnormality.

[0079] In the embodiment of the present application, it is also necessary to optimize and correct the component abnormality output by the abnormality identifier, and combine it with the actual frequency of occurrence of historical components, that is, the commonness of similar components in past buildings, so that the final component abnormality is more in line with the actual engineering scenario, avoiding misjudgment of abnormality due to the limitations of the identifier itself, improving the accuracy and reliability of abnormality quantification, and providing a more accurate basis for subsequent building component classification.

[0080] First, it is necessary to retrieve the historical component library, which stores a large number of historical building components from past buildings. Each component is associated with its morphological characteristic parameters such as curvature and dimensional deviation, and the frequency of occurrence. The frequency of occurrence is the number of times this type of component has been used in historical projects.

[0081] Next, using the first morphological characteristic parameter of the first building component as a benchmark, we search the historical component library for historical building components with similar parameters. We select multiple components with a high degree of matching morphological characteristic parameters and select them as first matching components, i.e., historical components with a similar morphology to the first building component. We then summarize the occurrence frequencies of all first matching components in the historical component library to obtain the first total occurrence frequency.

[0082] The first total occurrence frequency is compared with a preset basic occurrence frequency to calculate a first abnormality correction coefficient. The basic occurrence frequency is a set standard frequency value representing the commonness of a certain type of morphological component in conventional projects, such as 15 times.

[0083] The first abnormality correction coefficient may be calculated as follows: when the first total occurrence frequency is less than or equal to the basic occurrence frequency, the first abnormality correction coefficient is 1. When the first total occurrence frequency is greater than the basic occurrence frequency, the correction coefficient is calculated according to the formula "first abnormality correction coefficient = basic occurrence frequency / first total occurrence frequency".

[0084] Finally, the determined first abnormality correction coefficient is multiplied by the first component abnormality of the first building component. The result obtained is the corrected first component abnormality, which is used as the final component abnormality of the component.

[0085] For example, the anomaly degree identifier calculates the anomaly degree of a curtain wall component to be 0.6. After accessing the historical component library, three matching components were found that matched the morphological characteristic parameters of the first component. Their occurrence frequencies were 8, 6, and 7, respectively, for a total of 21 occurrences.

[0086] Assuming the preset base occurrence frequency is 15, since the total first occurrence frequency of 21 is greater than the base occurrence frequency of 15, the first anomaly correction coefficient is calculated according to the formula = 15 / 21 ≈ 0.71. The corrected first component anomaly is 0.6 × 0.71 ≈ 0.43, which is the final component anomaly for the first building component.

[0087] Finally, it is necessary to determine the first type of building components, the second type of building components and the third type of building components of the target special-shaped building based on the building geometric information.

[0088] In the embodiment of the present application, the core purpose of this step is to accurately classify all components of the target special-shaped building according to the complexity of the shape of the building components and the difficulty of modeling, to clarify which are conventional and easy-to-model components, which are complex components that need to be adapted to empirical modeling, and which are completely special-shaped components without precedent, so as to provide a classification basis for the subsequent differentiated setting of the point cloud scanning density, ensure that scanning resources are tilted towards high-difficulty components, and avoid excessive investment in simple components.

[0089] Specifically, it is necessary to classify the building components according to their abnormality levels calculated through the aforementioned steps.

[0090] When the abnormality of a component is lower than the first abnormality threshold, it is determined to be a first-class building component. This type of component is generally a conventional component and has a mature modeling method.

[0091] When the abnormality of a component is between the first abnormality threshold and the second abnormality threshold, it is determined to be a second-class building component. This type of component is generally a complex component and can be adaptively modeled based on existing experience.

[0092] When the abnormality of a component is higher than the second abnormality threshold, it is determined to be a third type of building component. This type of component is generally a complex component or a completely irregular component, and generally has no modeling precedent.

[0093] The first abnormality threshold is smaller than the second abnormality threshold. The specific abnormality threshold setting can be determined according to the actual situation in implementation. For example, the first abnormality threshold and the second abnormality threshold can be set to 0.3 and 0.7.

[0094] In step S200 of the embodiment of the present application, setting a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density, includes:

[0095] Receive standard component point cloud scanning density;

[0096] Setting the point cloud scanning density of the standard component to the first point cloud scanning density of the first type of building components;

[0097] According to the component abnormality of each building component in the second category of building components, a concentrated value of the abnormality of the second category of components is obtained, and combined with the point cloud scanning density of the standard component, a second point cloud scanning density is obtained;

[0098] According to the component abnormality of each building component in the third category of building components, the concentrated values ​​of the abnormality of the three categories of components are obtained, and combined with the point cloud scanning density of the standard component, the third point cloud scanning density is obtained.

[0099] The core purpose of step S200 in the embodiment of the present application is to match differentiated point cloud scanning densities for different categories of building components, so that the scanning resources can accurately adapt to the modeling requirements of the components. That is, a lower point cloud scanning density is used for the first category of building components with simple shapes and mature modeling; a medium point cloud scanning density is used for the second category of building components that require adaptation experience; and the highest point cloud scanning density is used for the third category of building components that are completely irregular. In this way, while ensuring the modeling accuracy of complex components, redundant data of simple building components is reduced, and the model quality and scanning efficiency are balanced.

[0100] First, you need to preset the "standard component point cloud scanning density", that is, the general scanning density for conventional standard components in the industry, as a benchmark value. For example, the preset standard component point cloud scanning density is 100 points / square meter, that is, 100 point cloud data are collected per square meter.

[0101] Then, it is necessary to set the scanning density of the first type of building components. In the specific implementation process, since the first type of building components are conventional components, there is no need to increase the density. Therefore, the point cloud scanning density of standard components can be directly determined as the "first point cloud scanning density" of the first type of building components, such as 100 points / square meter.

[0102] Next, the abnormality of all components in the second category of building components needs to be counted and the concentration value of the abnormality of the second category of components needs to be calculated. The concentration value of the abnormality of the second category of components can be an average value or other indicator that reflects the overall abnormality of the components in that category. For example, the concentration value of the abnormality of the second category of building components is 0.5. Optionally, the point cloud scan density can be calculated as follows: point cloud scan density = standard component point cloud scan density × (1 + second category component abnormality concentration value). Here, the second point cloud scan density corresponding to the second category of building components = 100 × (1 + 0.5) = 100 × 1.5 = 150 points / square meter.

[0103] Based on the same calculation logic as the second point cloud scan density, the third point cloud scan density can be set in combination with the component anomaly of each building component in the third category. For example, the third point cloud scan density = standard component point cloud scan density × (1 + the concentration value of the three-category component anomaly). The calculation logic for the third-category component anomaly concentration value is the same as the calculation logic for the second-category component anomaly concentration value. If the concentration value of the third-category component anomaly concentration value is 0.9, then the third point cloud scan density = 100 × (1 + 0.9) = 100 × 1.9 = 190 points / square meter.

[0104] The above method can be used to calculate the point cloud scan density of the three types of building structures. This allows point cloud scanning resources to be precisely matched to the scanning requirements of building components. This ensures high-precision data for the third type of completely irregular building components while avoiding data redundancy for the first type of conventional building components. This balances model accuracy and scanning efficiency, reducing data processing costs.

[0105] In step S300 of the embodiment of the present application, performing differentiated point cloud scanning on the target irregular-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud dataset includes:

[0106] Determining spatial position information of the first type of building components, the second type of building components, and the third type of building components in the target special-shaped building;

[0107] Dividing the target special-shaped building into a first scanning area, a second scanning area, and a third scanning area according to the spatial position information;

[0108] collecting point cloud data in the first scanning area according to the first point cloud scanning density to obtain a first point cloud data set;

[0109] collecting point cloud data in the second scanning area according to the second point cloud scanning density to obtain a second point cloud data set;

[0110] performing point cloud data collection in the third scanning area according to the third point cloud scanning density to obtain a third point cloud data set;

[0111] The first point cloud dataset, the second point cloud dataset, and the third point cloud dataset are aggregated to obtain the point cloud dataset.

[0112] In an embodiment of the present application, in order to perform differentiated point cloud scanning on the target special-shaped building and obtain a point cloud data set, it is first necessary to determine the spatial position information of the three types of building components in the target special-shaped building, such as the three-dimensional coordinate range, the floor or area where they are located, and then it is necessary to divide the target special-shaped building into a first scanning area, a second scanning area and a third scanning area according to the above spatial position information, which correspond to the distribution ranges of the three types of building components respectively.

[0113] Furthermore, it is necessary to collect data in the first scanning area according to the first point cloud scanning density to obtain a first point cloud data set; collect data in the second scanning area according to the second point cloud scanning density to obtain a second point cloud data set; and collect data in the third scanning area according to the third point cloud scanning density to obtain a third point cloud data set.

[0114] Finally, the first point cloud dataset, the second point cloud dataset, and the third point cloud dataset are integrated to form a point cloud dataset covering the entire target special-shaped building.

[0115] For example, in a theater, the first type of building components, such as rectangular auditorium seat frames, are distributed in the first-floor audience area. Their spatial locations are correspondingly divided into the first scanning area, and the first point cloud scanning density is 100 points / square meter to obtain the first point cloud dataset.

[0116] The second type of building components, such as curved side walls, are distributed in the second-floor corridor and are divided into the second scanning area. They are collected at a second point cloud scanning density of 150 points / square meter to obtain the second point cloud dataset.

[0117] The third type of building components, such as the twisted ceiling steel structure, are distributed on the top dome and are divided into the third scanning area. The third point cloud scanning density is 190 points / square meter, and the third point cloud dataset is obtained.

[0118] After aggregating the three datasets, a complete point cloud dataset of the theater is formed, which retains the complex details of the ceiling steel structure while avoiding data redundancy in the auditorium area.

[0119] In step S400 of the embodiment of the present application, a digital twin model of the target special-shaped building is constructed based on the point cloud dataset to obtain a target digital twin model.

[0120] The core purpose of step S400 in this embodiment is to construct a digital twin model that accurately reflects the morphological characteristics of the target irregular-shaped building based on the differentially collected point cloud dataset. Virtually mapping the building entity through point cloud data ensures both detailed restoration of the third-category building components and the integrity and accuracy of the overall model, ultimately resulting in a target digital twin model that can be used for full lifecycle management of the building.

[0121] First, the point cloud dataset obtained in step S300 needs to be processed by denoising, splicing, and registration to eliminate noise points and duplicate points that may exist in the data acquisition process and ensure the spatial consistency of point cloud data in different regions.

[0122] Then, based on the processed point cloud dataset, the point cloud data is converted into a 3D model using 3D modeling algorithms such as surface reconstruction and mesh generation. Specifically, for the areas corresponding to the third point cloud dataset, the focus is on restoring complex morphological details; for the areas corresponding to the first and second point cloud datasets, the model complexity is optimized while ensuring structural accuracy.

[0123] Finally, the three-dimensional models of each area are integrated into an overall model, the continuity and accuracy of the model are checked, and the joints are optimized and adjusted to finally obtain the target digital twin model of the target special-shaped building.

[0124] Example 2, as Figure 2 As shown, based on the same inventive concept as the method for constructing a digital twin model of a special-shaped building provided in Example 1, an embodiment of the present invention further provides a system for constructing a digital twin model of a special-shaped building, comprising:

[0125] A geometric information acquisition module 11 is configured to acquire architectural geometric information of a target special-shaped building, and determine first-category architectural components, second-category architectural components, and third-category architectural components of the target special-shaped building based on the architectural geometric information;

[0126] a point cloud density allocation module 12, configured to set a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density;

[0127] a point cloud scanning execution module 13, configured to perform a differentiated point cloud scan on the target irregular-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud data set;

[0128] The digital model construction module 14 is used to construct a digital twin model of the target special-shaped building according to the point cloud data set to obtain a target digital twin model.

[0129] Furthermore, the geometric information acquisition module 11 includes the following execution steps:

[0130] determining a plurality of building components according to the building geometric information, and extracting morphological characteristic parameters of each of the building components;

[0131] determining a component abnormality degree of each of the building components based on the morphological characteristic parameters of each of the building components;

[0132] Comparing the component abnormality of each of the building components with a preset first abnormality threshold and a second abnormality threshold, wherein the second abnormality threshold is greater than the first abnormality threshold;

[0133] When the component abnormality is less than the first abnormality threshold, classifying the corresponding building component as a first type of building component;

[0134] When the abnormality of a component is greater than or equal to the first abnormality threshold and less than the second abnormality threshold, classifying the corresponding building component as a second type of building component;

[0135] When the component abnormality is greater than or equal to the second abnormality threshold, the corresponding building component is classified as a third type of building component.

[0136] Wherein, determining the component abnormality of each building component based on the morphological characteristic parameters of each building component includes:

[0137] extracting a first building component from the plurality of building components, and obtaining a morphological characteristic parameter of the first building component as a first morphological characteristic parameter;

[0138] Retrieving an abnormality degree identifier, and processing the first morphological characteristic parameter by the abnormality degree identifier to obtain a first component abnormality degree of the first building component;

[0139] The component abnormality degrees of the remaining building components are obtained in the same manner as that of the first building component, thereby obtaining the component abnormality degree of each building component.

[0140] The steps of constructing the abnormality identifier include:

[0141] Collecting a plurality of sample building components, acquiring morphological characteristic parameters of the plurality of sample building components, and obtaining a sample morphological characteristic parameter set;

[0142] performing component abnormality identification on the plurality of sample building components according to the sample morphological characteristic parameter set to obtain a sample component abnormality degree set;

[0143] The abnormality identifier is trained and obtained based on the sample morphological feature parameter set and the sample component abnormality set using machine learning.

[0144] After obtaining the first component abnormality degree of the first building component, the method further includes:

[0145] Retrieving a historical component library, wherein the historical component library includes a plurality of historical building components and morphological characteristic parameters and occurrence frequencies of each of the historical building components;

[0146] Searching and matching the morphological characteristic parameter pairs of each of the historical building components in the historical component library based on the first morphological characteristic parameter to determine a plurality of first matching components;

[0147] Counting the occurrence frequencies of the plurality of first matching components to obtain a first total occurrence frequency;

[0148] determining a first abnormality correction coefficient according to the first total occurrence frequency and a preset basic occurrence frequency;

[0149] The first component abnormality is corrected using the first abnormality correction coefficient, and the correction result is used as the first component abnormality.

[0150] Furthermore, the point cloud density allocation module 12 includes the following execution steps:

[0151] Receive standard component point cloud scanning density;

[0152] Setting the point cloud scanning density of the standard component to the first point cloud scanning density of the first type of building components;

[0153] According to the component abnormality of each building component in the second category of building components, a concentrated value of the abnormality of the second category of components is obtained, and combined with the point cloud scanning density of the standard component, a second point cloud scanning density is obtained;

[0154] According to the component abnormality of each building component in the third category of building components, the concentrated values ​​of the abnormality of the three categories of components are obtained, and combined with the point cloud scanning density of the standard component, the third point cloud scanning density is obtained.

[0155] Furthermore, the point cloud scanning execution module 13 includes the following execution steps:

[0156] Determining spatial position information of the first type of building components, the second type of building components, and the third type of building components in the target special-shaped building;

[0157] Dividing the target special-shaped building into a first scanning area, a second scanning area, and a third scanning area according to the spatial position information;

[0158] collecting point cloud data in the first scanning area according to the first point cloud scanning density to obtain a first point cloud data set;

[0159] collecting point cloud data in the second scanning area according to the second point cloud scanning density to obtain a second point cloud data set;

[0160] performing point cloud data collection in the third scanning area according to the third point cloud scanning density to obtain a third point cloud data set;

[0161] The first point cloud dataset, the second point cloud dataset, and the third point cloud dataset are aggregated to obtain the point cloud dataset.

[0162] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0163] Those skilled in the art will appreciate that embodiments of the present invention may provide methods, systems, or computer program products. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0164] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0165] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0166] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0167] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0168] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for constructing a digital twin model of a special-shaped building, characterized in that: The method comprises: Acquiring architectural geometric information of a target special-shaped building, and determining first-category architectural components, second-category architectural components, and third-category architectural components of the target special-shaped building based on the architectural geometric information, including: determining a plurality of building components according to the building geometric information, and extracting morphological characteristic parameters of each of the building components; Determining the component abnormality of each building component based on the morphological characteristic parameters of each building component includes: extracting a first building component from the plurality of building components, and obtaining a morphological characteristic parameter of the first building component as a first morphological characteristic parameter; Retrieving an abnormality degree identifier, and processing the first morphological characteristic parameter by the abnormality degree identifier to obtain a first component abnormality degree of the first building component; Obtaining component abnormality degrees of the remaining building components in the same manner as that of the first component abnormality degree of the first building component, to obtain the component abnormality degree of each building component; Comparing the component abnormality of each of the building components with a preset first abnormality threshold and a second abnormality threshold, wherein the second abnormality threshold is greater than the first abnormality threshold; When the component abnormality is less than the first abnormality threshold, classifying the corresponding building component as a first type of building component; When the abnormality of a component is greater than or equal to the first abnormality threshold and less than the second abnormality threshold, classifying the corresponding building component as a second type of building component; When the abnormality degree of a component is greater than or equal to the second abnormality threshold, classifying the corresponding building component as a third type of building component; Setting a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density, comprises: Receive standard component point cloud scanning density; Setting the point cloud scanning density of the standard component to the first point cloud scanning density of the first type of building components; According to the component abnormality of each building component in the second category of building components, a concentrated value of the abnormality of the second category of components is obtained, and combined with the point cloud scanning density of the standard component, a second point cloud scanning density is obtained; According to the component abnormality of each building component in the third category of building components, a concentrated value of the abnormality of the three categories of components is obtained, and combined with the point cloud scanning density of the standard component, a third point cloud scanning density is obtained; Performing a differentiated point cloud scan on the target special-shaped building based on the first point cloud scan density, the second point cloud scan density, and the third point cloud scan density to obtain a point cloud dataset; Based on the point cloud data set, a digital twin model of the target special-shaped building is constructed to obtain a target digital twin model.

2. The method according to claim 1, characterized in that The steps of constructing the abnormality identifier include: Collecting a plurality of sample building components, acquiring morphological characteristic parameters of the plurality of sample building components, and obtaining a sample morphological characteristic parameter set; performing component abnormality identification on the plurality of sample building components according to the sample morphological characteristic parameter set to obtain a sample component abnormality degree set; The abnormality identifier is trained and obtained based on the sample morphological feature parameter set and the sample component abnormality set using machine learning.

3. The method according to claim 1, characterized in that After obtaining the first component abnormality degree of the first building component, the method further includes: Retrieving a historical component library, wherein the historical component library includes a plurality of historical building components and morphological characteristic parameters and occurrence frequencies of each of the historical building components; Searching and matching the morphological characteristic parameter pairs of each of the historical building components in the historical component library based on the first morphological characteristic parameter to determine a plurality of first matching components; Counting the occurrence frequencies of the plurality of first matching components to obtain a first total occurrence frequency; determining a first abnormality correction coefficient according to the first total occurrence frequency and a preset basic occurrence frequency; The first component abnormality is corrected using the first abnormality correction coefficient, and the correction result is used as the first component abnormality.

4. The method according to claim 1, wherein Performing a differentiated point cloud scan on the target irregular-shaped building based on the first point cloud scan density, the second point cloud scan density, and the third point cloud scan density to obtain a point cloud data set includes: Determining spatial position information of the first type of building components, the second type of building components, and the third type of building components in the target special-shaped building; Dividing the target special-shaped building into a first scanning area, a second scanning area, and a third scanning area according to the spatial position information; collecting point cloud data in the first scanning area according to the first point cloud scanning density to obtain a first point cloud data set; collecting point cloud data in the second scanning area according to the second point cloud scanning density to obtain a second point cloud data set; performing point cloud data collection in the third scanning area according to the third point cloud scanning density to obtain a third point cloud data set; The first point cloud dataset, the second point cloud dataset, and the third point cloud dataset are aggregated to obtain the point cloud dataset.

5. A digital twin model construction and processing system for special-shaped buildings, characterized by: The system comprises: A geometric information acquisition module is used to acquire architectural geometric information of a target special-shaped building, and determine the first type of architectural components, the second type of architectural components, and the third type of architectural components of the target special-shaped building based on the architectural geometric information, including: determining a plurality of building components according to the building geometric information, and extracting morphological characteristic parameters of each of the building components; Determining the component abnormality of each building component based on the morphological characteristic parameters of each building component includes: extracting a first building component from the plurality of building components, and obtaining a morphological characteristic parameter of the first building component as a first morphological characteristic parameter; Retrieving an abnormality degree identifier, and processing the first morphological characteristic parameter by the abnormality degree identifier to obtain a first component abnormality degree of the first building component; Obtaining component abnormality degrees of the remaining building components in the same manner as that of the first component abnormality degree of the first building component, to obtain the component abnormality degree of each building component; Comparing the component abnormality of each of the building components with a preset first abnormality threshold and a second abnormality threshold, wherein the second abnormality threshold is greater than the first abnormality threshold; When the component abnormality is less than the first abnormality threshold, classifying the corresponding building component as a first type of building component; When the abnormality of a component is greater than or equal to the first abnormality threshold and less than the second abnormality threshold, classifying the corresponding building component as a second type of building component; When the abnormality degree of a component is greater than or equal to the second abnormality threshold, classifying the corresponding building component as a third type of building component; a point cloud density allocation module, configured to set a first point cloud scanning density, a second point cloud scanning density, and a third point cloud scanning density according to the first type of building components, the second type of building components, and the third type of building components, wherein the third point cloud scanning density is greater than the second point cloud scanning density, and the second point cloud scanning density is greater than the first point cloud scanning density, comprising: Receive standard component point cloud scanning density; Setting the point cloud scanning density of the standard component to the first point cloud scanning density of the first type of building components; According to the component abnormality of each building component in the second category of building components, a concentrated value of the abnormality of the second category of components is obtained, and combined with the point cloud scanning density of the standard component, a second point cloud scanning density is obtained; According to the component abnormality of each building component in the third category of building components, a concentrated value of the abnormality of the three categories of components is obtained, and combined with the point cloud scanning density of the standard component, a third point cloud scanning density is obtained; a point cloud scanning execution module, configured to perform a differentiated point cloud scan on the target irregular-shaped building based on the first point cloud scanning density, the second point cloud scanning density, and the third point cloud scanning density to obtain a point cloud data set; The digital model construction module is used to construct a digital twin model of the target special-shaped building according to the point cloud data set to obtain a target digital twin model.

Images