Digital modeling method and platform for green fabricated building based on BIM technology

Abstract

This application relates to the field of data processing technology, specifically to a digital modeling method and platform for green prefabricated buildings based on BIM technology. The method includes: scanning the main structure of the building to be assembled at multiple pre-set monitoring stations to obtain point cloud data for each station; acquiring the feature difference values ​​of a single point cloud data point at a single monitoring station; constructing a spatial window for a single point cloud data point, acquiring the different degrees of the single point cloud data point, and classifying all point cloud data points; acquiring the filtering adjustment coefficients for each point cloud data point in each category, adjusting the minimum number of nearest neighbors when performing radius filtering on each point cloud data point in each category, and performing filtering processing on the point cloud data; and constructing a digital model of the green prefabricated building reflecting the actual construction status. This application aims to improve the accuracy of model geometric dimensions, the modeling accuracy of key assembly nodes, and the accuracy of green performance analysis by improving the processing accuracy of point cloud data.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a digital modeling method and platform for green prefabricated buildings based on BIM technology. Background Technology

[0002] As the construction industry transforms towards green, industrialized, and digital development, green prefabricated buildings have become the core of the industry's development, offering advantages such as factory production of prefabricated components, on-site modular assembly, low construction energy consumption, minimal construction waste, and resource recyclability. Digital modeling of green prefabricated buildings based on BIM technology enables full lifecycle management of buildings; for example, it supports multi-disciplinary collaboration during the design phase, avoiding issues such as component collisions and pipeline conflicts, and optimizing prefabricated component disassembly schemes; it provides precise assembly guidance during the construction phase, improving efficiency and quality; and it supports simulation and optimization of green performance aspects such as energy consumption, lighting, and ventilation during the operation and maintenance phase.

[0003] However, in the process of digitally modeling green prefabricated buildings based on BIM technology, the complex construction site environment, dust, equipment obstruction and light interference can easily introduce a large number of noise points. Traditional point cloud processing methods usually use globally uniform denoising parameters, which fail to distinguish the differences in geometric features between smooth component surfaces and complex assembly nodes. This makes it easy to misjudge real and complex assembly details as noise removal when filtering out noise, resulting in geometric size deviations in the BIM digital model, insufficient modeling accuracy of key assembly nodes and inaccurate green performance analysis, which cannot meet the needs of high-precision assembly and refined management. Summary of the Invention

[0004] In light of the above, it is necessary to provide a digital modeling method and platform for green prefabricated buildings based on BIM technology. Compared with traditional BIM-based digital modeling methods for green prefabricated buildings, this method improves the accuracy of model geometric dimensions, the modeling accuracy of key assembly nodes, and the accuracy of green performance analysis by enhancing the processing accuracy of point cloud data. In a first aspect, embodiments of this application provide a digital modeling method for green prefabricated buildings based on BIM technology, the method comprising the following steps: The main structure of the building to be assembled is scanned at multiple preset stations to obtain point cloud data for each station; For a single monitoring station, the feature difference value of a single point cloud data is obtained by comparing the local feature differences between a single point cloud data at the station and its preset nearest neighbor point cloud data. Based on the spatial distance between a single point cloud data and its nearest neighbor point cloud data, a spatial window for the single point cloud data is constructed. By comparing the distribution probability of the feature difference values ​​of the single point cloud data within its spatial window with those within its preset nearest neighbor spatial windows, and by comparing the total number of point cloud data within each nearest neighbor spatial window with the total number of point cloud data within all nearest neighbor spatial windows, the dissimilarity of the single point cloud data is obtained, thus classifying all point cloud data. Through the distribution of the dissimilarity of point cloud data in each category, the filtering adjustment coefficient of each point cloud data in each category is obtained, thereby adjusting the minimum number of nearest neighbor points when performing radius filtering on each point cloud data in each category, and thus performing filtering processing on the point cloud data. The point cloud data of all stations after filtering are spatially registered and fused. Based on the fusion results, the design basis data of the prefabricated building to be assembled are geometrically calibrated. Combined with the preset green performance attributes, a digital model of green prefabricated building reflecting the actual construction status is constructed.

[0005] In one embodiment, the process of obtaining the feature difference value is as follows: Calculate the similarity of the normal vectors between a single point cloud data point and its nearest neighbor point cloud data points; Calculate the difference in curvature between a single point cloud data point and its nearest neighbor point cloud data points; The feature difference value of a single point cloud data is obtained by measuring the similarity of the normal vectors between a single point cloud data and all its nearest neighbor point cloud data, as well as the difference in curvature between a single point cloud data and all its nearest neighbor point cloud data.

[0006] In one embodiment, the calculation process of the feature difference value is as follows: Calculate the average similarity of the normal vectors between a single point cloud data point and all its nearest neighbor point cloud data points; Calculate the arithmetic mean of the differences in curvature between a single point cloud data point and all its nearest neighbor point cloud data points; The characteristic difference value is positively correlated with the arithmetic mean and negatively correlated with the average value.

[0007] In one embodiment, the method for constructing the spatial window is as follows: Find the maximum Euclidean distance between a single point cloud data point and all its nearest neighbor point cloud data points; The spatial window is centered on a single point cloud data point, with the maximum value as its side length, and the sides of the spatial window are aligned with the axes of the spatial coordinate system.

[0008] In one embodiment, the process of obtaining the different degrees is as follows: The range of feature differences in point cloud data within a single point cloud data spatial window and all its nearest spatial windows is statistically analyzed, and the range of values ​​is divided into multiple equidistant intervals. Calculate the probability that the feature difference values ​​of point cloud data within each spatial window fall into each interval; Calculate the combined difference in probability between a spatial window of a single point cloud data point and its nearest neighbor spatial windows falling within all intervals; Calculate the proportion of point cloud data within each nearest neighbor spatial window to the total number of point cloud data within all nearest neighbor spatial windows; The difference is obtained by the combined difference between a spatial window of a single point cloud data and all its neighboring spatial windows, and the proportion of the number of all neighboring spatial windows.

[0009] In one embodiment, the calculation process for the different degrees is as follows: The product of the overall gap and the quantity ratio is denoted as the reliable product; The difference is the sum of the reliable products between a spatial window of a single point cloud data and all its nearest spatial windows.

[0010] In one embodiment, the filter adjustment coefficient is a normalized value of the mean of different degrees of all point cloud data in each class.

[0011] In one embodiment, the adjustment process for the minimum nearest neighbor number is as follows: The product of the preset maximum adjustment amount of the minimum nearest neighbor number and the filter adjustment coefficient is recorded as the actual adjustment amount; The actual value of the minimum nearest neighbor number is adjusted by the actual adjustment amount.

[0012] In one embodiment, the method for calculating the minimum number of nearest neighbors is as follows: The rounded result of the difference between the preset maximum value of the minimum nearest neighbor number and the actual adjustment amount is taken as the actual value of the minimum nearest neighbor number.

[0013] Secondly, embodiments of this application also provide a digital modeling platform for green prefabricated buildings based on BIM technology, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described digital modeling methods for green prefabricated buildings based on BIM technology.

[0014] This application has at least the following beneficial effects: This application obtains feature difference values ​​by comparing the local features of a single point cloud data with its preset neighboring point cloud data. This can quantify the feature differences between the location of each point cloud data and its neighboring locations, providing a basis for subsequent judgment of the geometric complexity of the location of the point cloud data. Furthermore, by constructing a spatial window based on the spatial distance between a single point cloud data point and its nearest neighbor point cloud data point, the size of the spatial window can be dynamically adjusted according to the point cloud distribution characteristics, avoiding the problem of an excessively large spatial window introducing irrelevant noise or an excessively small spatial window losing key structural information. By constructing a spatial window, a unified spatial analysis unit is provided for subsequent comparative analysis of the feature distribution patterns of a local area and its nearest neighbor areas, making the judgment of geometric complexity more accurate and reliable. By comparing the distribution probability of feature differences within the spatial window of a single point cloud data point with its preset nearest neighbor spatial windows, it is helpful to distinguish between the true geometric complexity and the additional complexity caused by noise. By combining the comparison results of the total number of point cloud data points within each nearest neighbor spatial window with the total number of point cloud data points within all nearest neighbor spatial windows, the reliability of the comparison results can be improved. The different degrees can accurately reflect the geometric complexity of the location of a single point cloud data point, avoiding confusion between true structural details and noise features, and helping to use a lighter filtering intensity for areas with high geometric complexity. Furthermore, by obtaining the filtering adjustment coefficients through the distribution of different degrees of point cloud data in various types, the geometric complexity of the location of point cloud data in various types can be adaptively reflected. Based on the filtering adjustment coefficients, the minimum number of nearest neighbor points when adjusting the radius filtering is balanced with the preservation of details of complex assembly nodes and the suppression of noise of simple assembly nodes, which significantly improves the processing accuracy of point cloud data, effectively matches the high-precision requirements of BIM modeling, improves the accuracy of model geometric dimensions, the modeling accuracy of key assembly nodes, and the accuracy of green performance analysis, and meets the needs of high-precision assembly and refined management of green prefabricated buildings based on digital models. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the steps of a BIM-based digital modeling method for green prefabricated buildings, as provided in one embodiment of this application; Figure 2 This is a schematic diagram of the filtering process for point cloud data. Detailed Implementation

[0017] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0019] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0020] The following, in conjunction with the accompanying drawings, details the specific scheme of the digital modeling method and platform for green prefabricated buildings based on BIM technology provided in this application.

[0021] Please see Figure 1 The document illustrates a flowchart of a BIM-based digital modeling method for green prefabricated buildings, according to an embodiment of this application. The method includes the following steps: Step 1: Scan the main structure of the building to be assembled at multiple preset stations to obtain point cloud data for each station.

[0022] To achieve high-precision digital modeling of green prefabricated buildings, a laser scanner is needed to perform a comprehensive scan of the building. Specifically: Multiple measuring stations are evenly deployed in the vicinity of the main structure of the building to be assembled to ensure comprehensive point cloud data acquisition of the main structure. The overlap area between adjacent measuring stations is no less than 30% to meet the requirements of subsequent data stitching. A unified coordinate origin and elevation datum are set based on the building construction control points. All 3D laser scanners are connected to the same spatial coordinate system, and coordinate calibration is performed using control points marked by a total station to ensure spatial consistency of multi-source data, laying the foundation for subsequent data stitching and modeling.

[0023] In this embodiment, before point cloud data is collected from the building to be assembled, the main vertical load-bearing components such as precast columns and the horizontal load-bearing components such as precast beams of the building to be assembled have been positioned, spliced ​​and temporarily fixed according to the design requirements, and the overall axis, elevation and key assembly nodes of the building have been formed.

[0024] In this embodiment, the number of stations is 8, the Euclidean distance between adjacent stations is 15m, the overlap area between adjacent stations is 30%, and the sampling interval of the 3D laser scanner is 0.5mm. The number of stations, the Euclidean distance between adjacent stations, the overlap area between adjacent stations, and the sampling interval are all preset by humans. The implementer can set them according to the actual situation. This application does not impose any special restrictions.

[0025] Step 2: For a single station, obtain the feature difference value of a single point cloud data; construct a spatial window for a single point cloud data, obtain the degree of difference of a single point cloud data, and classify all point cloud data; obtain the filtering adjustment coefficient of each point cloud data in each category, adjust the minimum number of nearest neighbors when performing radius filtering on each point cloud data in each category, and perform filtering processing on the point cloud data.

[0026] Because green prefabricated buildings adopt a modular construction method using prefabricated components, there may be significant structural deviations between different assembly nodes. This leads to variations in the impact of factors such as dust, equipment obstruction, and light interference on the positions of different assembly nodes in the actual construction environment. If a uniform denoising parameter is used to process point cloud data from different stations, it is difficult to balance the preservation of details for complex assembly nodes with the suppression of noise for simple assembly nodes, resulting in insufficient accuracy in point cloud data processing and affecting the subsequent detailed digital modeling of green prefabricated buildings.

[0027] Based on the above analysis, it is necessary to first quantify the local geometric features of each point cloud data location, then determine the geometric complexity of the point cloud data location based on the differences in local features, and finally cluster and divide regions with different geometric complexities and adaptively adjust the filtering parameters to improve the processing accuracy of point cloud data and avoid over-filtering of complex assembly nodes, which would reduce the accuracy of subsequent digital modeling of green prefabricated buildings.

[0028] Step 2.1: For a single station, obtain the feature difference value of the single point cloud data by the local feature difference between the single point cloud data of the single station and its preset neighboring point cloud data.

[0029] To distinguish between "real complex structures" and "environmental noise" in the future, it is necessary to quantify the local geometric features of each point cloud data location.

[0030] Based on the above analysis, for a single monitoring station, the feature difference value of a single point cloud data is obtained by comparing the local feature differences between a single point cloud data of a single monitoring station and its preset nearest neighbor point cloud data. The specific process is as follows: Calculate the similarity of the normal vectors between a single point cloud data point and its nearest neighbor point cloud data points; Calculate the difference in curvature between a single point cloud data point and its nearest neighbor point cloud data points; Calculate the average similarity of the normal vectors between a single point cloud data point and all its nearest neighbor point cloud data points; Calculate the arithmetic mean of the differences in curvature between a single point cloud data point and all its nearest neighbor point cloud data points; The feature difference value of a single point cloud data is positively correlated with the arithmetic mean and negatively correlated with the average value.

[0031] It should be noted that: positive correlation means that the variables change in the same direction, that is, when one variable increases, the other variable also increases, and when one variable decreases, the other variable also decreases; negative correlation means that the variables change in opposite directions, that is, when one variable increases, the other variable decreases, and when one variable decreases, the other variable increases.

[0032] In this embodiment, the nearest point cloud data of a single point cloud data is obtained through the K-nearest neighbor algorithm, where K is 100. The value of K is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions. The K-nearest neighbor algorithm is a well-known technology and will not be described in detail here.

[0033] In this embodiment, a single point cloud data and all its nearest neighbor point cloud data are used as input. The least squares method is used to output a fitted surface. The normal vector and curvature of the fitted surface are used as the normal vector and curvature of the single point cloud data. The surface fitting using the least squares method and the calculation of the normal vector and curvature are well-known techniques and will not be described in detail in this application.

[0034] In this embodiment, the similarity between normal vectors is cosine similarity. The calculation of cosine similarity is a well-known technique and will not be described in detail here. As other implementation methods, implementers may adopt other existing feasible techniques based on the ability to measure the similarity between normal vectors. This application does not impose any special restrictions.

[0035] In this embodiment, the difference between curvatures is the absolute value of the difference. As other implementation methods, based on the ability to measure the degree of difference between curvatures, the implementer may adopt other existing feasible technologies, such as the square of the difference, the ratio, etc. This application does not impose any special restrictions.

[0036] In this embodiment, the expression for the feature difference value of a single point cloud data is: In the formula, d represents the feature difference value of a single point cloud data. This represents the average similarity of the normal vectors between a single point cloud data point and all its nearest neighbor point cloud data points; This represents the arithmetic mean of the differences in curvature between a single point cloud data point and all its nearest neighbor point cloud data points; exp() represents an exponential function with the natural constant as the base, used to... The mapping is to positive numbers and normalized. The natural constant is merely one embodiment of this application; implementers may replace the natural constant with other values ​​greater than 1 according to actual circumstances, and this application does not impose any special restrictions.

[0037] It should be noted that the calculated average value... The larger the value, the more significant the orientation consistency of the local surface; the calculated arithmetic mean The larger the value, the more drastic the fluctuations in the local structure may be; the larger the calculated feature difference value, the greater the feature difference between the location of a single point cloud data point and its nearest neighbor.

[0038] Step 2.2: Based on the spatial distance between a single point cloud data and its neighboring point cloud data, construct a spatial window for the single point cloud data. By comparing the distribution probability of the feature difference values ​​of the single point cloud data within its spatial window with those of the point cloud data within its preset neighboring spatial windows, and by comparing the total number of point cloud data within each neighboring spatial window with the total number of point cloud data within all neighboring spatial windows, obtain the degree of difference of the single point cloud data.

[0039] Feature difference values ​​quantify the degree of feature difference between the location of a single point cloud data point and its nearest neighbors, but they cannot distinguish whether the location of a single point cloud data point represents a genuine geometrically complex structure or additional complexity introduced by noise points. Therefore, to differentiate between these two cases, it is necessary to consider the assembly characteristics of modular assembly. Complex assembly nodes are usually composed of multiple prefabricated components. Thus, the feature difference values ​​between a local region centered on a single point cloud data point and its nearest neighbors exhibit a certain spatial distribution pattern. If the feature distribution of a local region centered on a single point cloud data point differs significantly from that of its nearest neighbors, it indicates that the location of the single point cloud data point is more likely to have geometrically complex features.

[0040] Based on the above analysis, for a single monitoring station, a spatial window for a single point cloud data is constructed based on the spatial distance between a single point cloud data and its nearest neighbor point cloud data. The construction method is as follows: Find the maximum Euclidean distance between a single point cloud data point and all its nearest neighbor point cloud data points; The spatial window for a single point cloud data is centered on the single point cloud data, with the maximum value as the side length, and the side of the spatial window is aligned with the axis of the spatial coordinate system.

[0041] It should be noted that a spatial window is constructed by using the spatial distance between a single point cloud data and its nearest neighbor point cloud data. This allows the size of the spatial window to be dynamically adjusted according to the distribution characteristics of the point cloud, avoiding the problem of an excessively large spatial window introducing irrelevant noise or an excessively small spatial window losing key structural information.

[0042] Furthermore, for a single measurement station, the dissimilarity of a single point cloud data point is obtained by comparing the distribution probability of feature differences between the point cloud data within a spatial window and the point cloud data within a preset neighboring spatial window, and by comparing the total number of point cloud data points within each neighboring spatial window with the total number of point cloud data points within all neighboring spatial windows. The specific process is as follows: The range of feature differences in point cloud data within a single point cloud data spatial window and all its nearest spatial windows is statistically analyzed, and the range of values ​​is divided into multiple equidistant intervals. Calculate the probability that the feature difference values ​​of point cloud data within each spatial window fall into each interval; Calculate the combined difference in probability between a spatial window of a single point cloud data point and its nearest neighbor spatial windows falling within all intervals; Calculate the proportion of point cloud data within each nearest neighbor spatial window to the total number of point cloud data within all nearest neighbor spatial windows; The product of the overall gap and the quantity ratio is denoted as the reliable product; The sum of the reliable products between a spatial window of a single point cloud data and all its nearest neighbor spatial windows is taken as the distinctness of the single point cloud data.

[0043] In this embodiment, the nearest spatial windows of a single point cloud data spatial window are specifically: spatial windows that are connected to the six faces of the single point cloud data spatial window. These six spatial windows have the same size as the single point cloud data spatial window, and each of these six spatial windows is connected to the single point cloud data spatial window with one and only one face. If there are fewer than six nearest spatial windows, only the existing nearest spatial windows are used for subsequent calculations.

[0044] In this embodiment, the number of equidistant intervals is 10. The number of equidistant intervals is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0045] In this embodiment, the calculation process of the comprehensive gap is as follows: the probabilities of each spatial window falling into all intervals are arranged in interval order to form a discrete probability distribution sequence for each spatial window; the sum of the two-way KL divergence values ​​of the discrete probability distribution sequences between each spatial window and its nearest neighbor spatial windows is calculated, and this sum is used as the comprehensive gap. The calculation of the KL divergence value is a well-known technique and will not be elaborated upon in this application. As other implementation methods, implementers may use other existing feasible techniques to measure the degree of difference between discrete probability distribution sequences, and this application does not impose any special restrictions.

[0046] It should be noted that: the larger the calculated proportion, the higher the accuracy of the feature difference comparison results between the spatial window of a single point cloud data and its corresponding nearest spatial window; the larger the calculated difference, the more likely the location of a single point cloud data has geometrically complex features.

[0047] Step 2.3: Classify all point cloud data; by analyzing the distribution of different degrees of point cloud data in each category, obtain the filtering adjustment coefficients for each point cloud data in each category, so as to adjust the minimum number of nearest neighbors when performing radius filtering on each point cloud data in each category, thereby performing filtering processing on the point cloud data.

[0048] Because the structural complexity of different assembly nodes in prefabricated buildings varies, areas with more complex local structures are more likely to be confused with noise points under the influence of interference and be misjudged as noise points and removed. Therefore, it is necessary to classify point cloud data with different interference response characteristics, lightly filter complex assembly node areas to retain details, and heavily filter simple areas to suppress noise.

[0049] Based on the above analysis, the point cloud data of a single station are classified according to the different degrees of all point cloud data of a single station.

[0050] In this embodiment, agglomerative hierarchical clustering algorithm is used to classify all point cloud data of a single station. The metric distance is the absolute value of the difference between different degrees of point cloud data. The inter-cluster merging criterion adopts the Ward variance minimization criterion. The number of clusters is set to 4, which is calculated from experimental data. Both the agglomerative hierarchical clustering algorithm and the Ward variance minimization criterion are well-known technologies and will not be described in detail in this application. As other implementation methods, based on the ability to classify point cloud data according to different degrees, implementers can adopt other existing feasible technologies, which will not be described in detail in this application.

[0051] Furthermore, for a single station, by analyzing the distribution of different degrees of point cloud data across various categories, filtering adjustment coefficients for each type of point cloud data are obtained. These coefficients are used to adjust the minimum number of nearest neighbors required for radius filtering of each type of point cloud data, thereby filtering the point cloud data. The specific process is as follows: The normalized value of the mean of different degrees of all point cloud data in each category is used as the filtering adjustment coefficient of each point cloud data in each category. The product of the preset maximum adjustment amount of the minimum nearest neighbor number and the filter adjustment coefficient is recorded as the actual adjustment amount; The rounded result of the difference between the preset maximum value of the minimum nearest neighbor number and the actual adjustment amount is taken as the actual value of the minimum nearest neighbor number.

[0052] In this embodiment, the average of different degrees of all point cloud data in each class is recorded as the interference mean of each class. The normalized value of the interference mean of each class is obtained by the maximum value normalization method. In the normalization process, the maximum value refers to the maximum value among the interference means of all classes. The maximum value normalization method is a well-known technology and will not be described in detail in this application.

[0053] In this embodiment, the preset maximum adjustment amount is 14 and the preset maximum is 20. Both the preset maximum adjustment amount and the preset maximum value are calculated from experimental data.

[0054] It should be added that if the mean interference of all classes is equal, the filtering adjustment coefficient of the point cloud data in all classes should be uniformly set to 0.5.

[0055] It should be noted that the larger the calculated filter adjustment coefficient, the more likely the point cloud data in each category is to have geometrically complex features. Therefore, the number of minimum nearest neighbors should be reduced to preserve structural details and avoid increasing the deviation of the digital modeling structure of green prefabricated buildings due to excessive filtering.

[0056] Furthermore, a radius filtering algorithm is employed to filter the point cloud data from a single station, obtaining high-quality point cloud data for that station. This improves the processing accuracy of point clouds at complex structures in green prefabricated buildings, thereby enhancing the accuracy of subsequent model construction. The search radius of the radius filtering algorithm is 10 mm, calculated from experimental data. A schematic diagram of the point cloud data filtering process is shown below. Figure 2 As shown.

[0057] Step 3: Spatial registration and fusion of the point cloud data from all stations after filtering; geometric calibration of the design basis data of the prefabricated building to be assembled based on the fusion results; and construction of a digital model of the green prefabricated building that reflects the actual construction status by combining the preset green performance attributes.

[0058] To further construct accurate digital models of green prefabricated buildings based on BIM technology, this application obtains the basic design data of the building to be assembled. This basic design data includes architectural construction drawings, detailed design drawings of prefabricated components, component material parameters, and node connection process documents. The architectural construction drawings and detailed design drawings of prefabricated components are converted to DWG format, and the component material parameters are compiled into an Excel spreadsheet to ensure they can be accessed at any time during the modeling process. Furthermore, the application obtains green performance-related data of the building to be assembled, including environmental protection building material indicators, energy-saving coefficients, and sound and heat insulation parameters of the prefabricated components. It should be noted that the basic design data and green performance-related data are fundamental data for architectural design, obtained through publicly available documents. The specific acquisition and processing procedures are well-known to those skilled in the art and will not be elaborated upon here.

[0059] Furthermore, based on the high-quality point cloud data of each monitoring station obtained after filtering, and combined with the basic design data of the building to be assembled and the green performance correlation data, a digital model of the green prefabricated building is constructed. The specific process is as follows: Using the coordinate origin and elevation datum uniformly marked by the total station, the high-quality point cloud data of all stations are spatially aligned and redundantly removed, and then stitched together to generate complete high-quality point cloud data of the main structure of the building to be assembled, covering the entire construction scene, as a three-dimensional digital model reflecting the true geometric shape of the current construction status. In Revit software, the project grid and elevation system are first created based on the architectural construction drawings and the detailed design drawings of prefabricated components, defining the basic structural layout and component parameters in architectural theory. Then, the complete high-quality point cloud data of the main structure of the building to be assembled is used as the measured geometric reference layer and overlaid and matched with the theoretical design framework in the BIM model. Using the complete high-quality point cloud data, the complex assembly node features and overall structural outline are accurately preserved, and the geometric dimensions, spatial positions and assembly joints of the theoretical component model generated by Revit software are reverse-calibrated. By correcting the theoretical coordinates and tilt deviations of the prefabricated components, the idealized design drawings are transformed into a digital geometric model that perfectly matches the current actual construction status. At the same time, the green performance data is linked to the corresponding component and assembly node models, and the corrected geometric model is given specific environmental protection building material indicators, energy saving coefficients and sound and heat insulation parameters, etc., to achieve the integration of geometric model and green information. After modeling is completed, the constructed model is overlaid and compared with the complete, high-quality point cloud data of the main building structure to be assembled in Revit software using the point cloud comparison module. The geometric deviation between the model and the point cloud is calculated, and it is verified whether the geometric deviation meets the preset modeling accuracy requirements. If the accuracy requirements are met, a multi-format BIM model file containing geometric deviation and green performance attributes, along with a corresponding analysis report, is exported to accurately guide the subsequent assembly construction of the structure and provide high-precision digital support for the refined operation and maintenance management of the building throughout its entire life cycle. If the requirements are not met, the point cloud data for the out-of-tolerance areas is re-collected and processed until the preset modeling accuracy requirements are met. In this embodiment, the preset modeling accuracy requirement is that the geometric deviation is no more than 1.5 times the point cloud sampling interval. Implementers can set this requirement according to their actual situation, and this application does not impose any special restrictions.

[0060] Based on the same inventive concept as the above methods, this application also provides a digital modeling platform for green prefabricated buildings based on BIM technology, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described digital modeling methods for green prefabricated buildings based on BIM technology.

[0061] In summary, this application obtains feature difference values ​​by comparing the local features of a single point cloud data with its preset neighboring point cloud data. This can quantify the feature differences between the location of each point cloud data and its neighboring locations, providing a basis for subsequent judgment of the geometric complexity of the location of the point cloud data. Furthermore, by constructing a spatial window based on the spatial distance between a single point cloud data point and its nearest neighbor point cloud data point, the size of the spatial window can be dynamically adjusted according to the point cloud distribution characteristics, avoiding the problem of an excessively large spatial window introducing irrelevant noise or an excessively small spatial window losing key structural information. By constructing a spatial window, a unified spatial analysis unit is provided for subsequent comparative analysis of the feature distribution patterns of a local area and its nearest neighbor areas, making the judgment of geometric complexity more accurate and reliable. By comparing the distribution probability of feature differences within the spatial window of a single point cloud data point with its preset nearest neighbor spatial windows, it is helpful to distinguish between the true geometric complexity and the additional complexity caused by noise. By combining the comparison results of the total number of point cloud data points within each nearest neighbor spatial window with the total number of point cloud data points within all nearest neighbor spatial windows, the reliability of the comparison results can be improved. The different degrees can accurately reflect the geometric complexity of the location of a single point cloud data point, avoiding confusion between true structural details and noise features, and helping to use a lighter filtering intensity for areas with high geometric complexity. Furthermore, by obtaining the filtering adjustment coefficients through the distribution of different degrees of point cloud data in various types, the geometric complexity of the location of point cloud data in various types can be adaptively reflected. Based on the filtering adjustment coefficients, the minimum number of nearest neighbor points when adjusting the radius filtering is balanced with the preservation of details of complex assembly nodes and the suppression of noise of simple assembly nodes, which significantly improves the processing accuracy of point cloud data, effectively matches the high-precision requirements of BIM modeling, improves the accuracy of model geometric dimensions, the modeling accuracy of key assembly nodes, and the accuracy of green performance analysis, and meets the needs of high-precision assembly and refined management of green prefabricated buildings based on digital models.

[0062] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0063] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. A digital modeling method for green prefabricated buildings based on BIM technology, characterized in that, The method includes the following steps: The main structure of the building to be assembled is scanned at multiple preset stations to obtain point cloud data for each station; For a single monitoring station, the feature difference value of a single point cloud data is obtained by comparing the local feature differences between a single point cloud data at the station and its preset nearest neighbor point cloud data. Based on the spatial distance between a single point cloud data and its nearest neighbor point cloud data, a spatial window for the single point cloud data is constructed. By comparing the distribution probability of the feature difference values ​​of the single point cloud data within its spatial window with those within its preset nearest neighbor spatial windows, and by comparing the total number of point cloud data within each nearest neighbor spatial window with the total number of point cloud data within all nearest neighbor spatial windows, the dissimilarity of the single point cloud data is obtained, thus classifying all point cloud data. Through the distribution of the dissimilarity of point cloud data in each category, the filtering adjustment coefficient of each point cloud data in each category is obtained, thereby adjusting the minimum number of nearest neighbor points when performing radius filtering on each point cloud data in each category, and thus performing filtering processing on the point cloud data. The point cloud data of all stations after filtering are spatially registered and fused. Based on the fusion results, the design basis data of the prefabricated building to be assembled are geometrically calibrated. Combined with the preset green performance attributes, a digital model of green prefabricated building reflecting the actual construction status is constructed. The method for constructing the spatial window is as follows: Find the maximum Euclidean distance between a single point cloud data point and all its nearest neighbor point cloud data points; The spatial window is centered on a single point cloud data point, with the maximum value as its side length, and the sides of the spatial window are aligned with the axes of the spatial coordinate system.

2. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 1, characterized in that, The process for obtaining the feature difference values ​​is as follows: Calculate the similarity of the normal vectors between a single point cloud data point and its nearest neighbor point cloud data points; Calculate the difference in curvature between a single point cloud data point and its nearest neighbor point cloud data points; The feature difference value of a single point cloud data is obtained by measuring the similarity of the normal vectors between a single point cloud data and all its nearest neighbor point cloud data, as well as the difference in curvature between a single point cloud data and all its nearest neighbor point cloud data.

3. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 2, characterized in that, The calculation process for the feature difference value is as follows: Calculate the average similarity of the normal vectors between a single point cloud data point and all its nearest neighbor point cloud data points; Calculate the arithmetic mean of the differences in curvature between a single point cloud data point and all its nearest neighbor point cloud data points; The characteristic difference value is positively correlated with the arithmetic mean and negatively correlated with the average value.

4. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 1, characterized in that, The process of obtaining the different degrees is as follows: The range of feature differences in point cloud data within a single point cloud data spatial window and all its nearest spatial windows is statistically analyzed, and the range of values ​​is divided into multiple equidistant intervals. Calculate the probability that the feature difference values ​​of point cloud data within each spatial window fall into each interval; Calculate the combined difference in probability between a spatial window of a single point cloud data point and its nearest neighbor spatial windows falling within all intervals; Calculate the proportion of point cloud data within each nearest neighbor spatial window to the total number of point cloud data within all nearest neighbor spatial windows; The difference is obtained by the combined difference between a spatial window of a single point cloud data and all its neighboring spatial windows, and the proportion of the number of all neighboring spatial windows.

5. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 4, characterized in that, The calculation process for the different degrees is as follows: The product of the overall gap and the quantity ratio is denoted as the reliable product; The difference is the sum of the reliable products between a spatial window of a single point cloud data and all its nearest spatial windows.

6. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 1, characterized in that, The filter adjustment coefficient is the normalized value of the mean of different degrees of all point cloud data in each category.

7. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 1, characterized in that, The process for adjusting the minimum number of nearest neighbors is as follows: The product of the preset maximum adjustment amount of the minimum nearest neighbor number and the filter adjustment coefficient is recorded as the actual adjustment amount; The actual value of the minimum nearest neighbor number is adjusted by the actual adjustment amount.

8. The digital modeling method for green prefabricated buildings based on BIM technology as described in claim 7, characterized in that, The method for calculating the minimum number of nearest neighbors is as follows: The rounded result of the difference between the preset maximum value of the minimum nearest neighbor number and the actual adjustment amount is taken as the actual value of the minimum nearest neighbor number.

9. A digital modeling platform for green prefabricated buildings based on BIM technology, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the digital modeling method for green prefabricated buildings based on BIM technology as described in any one of claims 1-8.

Images