A bim-based support and hanger design method and system

Abstract

The application discloses a kind of based on BIM's support hanger design method and system, it is related to building engineering design technical field.The method includes: obtaining the environment BIM model including main structure BIM data and pipeline BIM data, constraint setting is carried out, obtains support hanger design constraint and the constraint strength of characterizing its strictness;Based on pipeline BIM data and main structure BIM data, random design of support hanger is carried out, and first support hanger design scheme is generated;According to environment BIM model, fitness analysis calculation is carried out to first support hanger design scheme, and first support hanger fitness is obtained;Optimization cluster mechanism is set based on constraint strength and scheme optimization is carried out, and optimal support hanger scheme is obtained as design result.The application realizes the automation and intelligentization of support hanger design, and optimal arrangement scheme meeting safety, economy requirement can be efficiently generated.

Description

Technical Field

[0001] This invention relates to the field of architectural engineering design technology, specifically to a BIM-based support and hanger design method and system. Background Technology

[0002] BIM, or Building Information Modeling, is a digital tool used in engineering design, construction, and management. It integrates and manages various physical and functional information of a building project through a three-dimensional model. Pipe supports and hangers are important structural components used to support and fix various pipes, cables, and other facilities. Their design must comprehensively consider multiple factors such as the building's main structural conditions, pipeline spatial layout, load distribution, and economic efficiency, directly affecting the safety, stability, and service life of the pipeline system.

[0003] Existing support and hanger design methods and software tools generally suffer from insufficient intelligence and limited optimization capabilities. On the one hand, current technologies struggle to automatically extract precise support requirements and constraint strengths from BIM data, and the generation of support and hanger design schemes often relies on fixed rules, resulting in insufficient diversity and difficulty in finding optimal solutions. On the other hand, scheme evaluation often neglects accurate prediction of overall stability and lacks a comprehensive balance between stability and economy; the optimization process is rigid and cannot adaptively adjust strategies based on constraint strength, leading to problems such as insufficient stability or excessive material usage. Furthermore, it is difficult to converge under complex conditions, failing to meet actual engineering requirements. Summary of the Invention

[0004] This invention addresses the technical problems of low automation in existing support and hanger design, difficulty in achieving multi-objective intelligent optimization, and inability to dynamically adjust design strategies according to engineering constraints. It provides a BIM-based support and hanger design method and system.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] In a first aspect, the present invention provides a BIM-based support and hanger design method, comprising:

[0007] Obtain the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths. The environmental BIM model includes the main structure BIM data and the piping BIM data.

[0008] Based on the pipeline BIM data and the main structure BIM data, the supports and hangers are randomly designed under the constraints of the support and hanger design to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers.

[0009] Based on the environmental BIM model, the fitness analysis calculation of the first hanger design scheme is performed to obtain the fitness of the first hanger. The fitness calculation includes predicting the first stability and the first material usage of the first hanger design scheme.

[0010] The design scheme of the support and hanger is optimized to obtain the optimal support and hanger scheme as the design result. In this process, an optimization cluster mechanism is set up based on the constraint strength to perform optimization.

[0011] Secondly, the present invention provides a BIM-based support and hanger design system, comprising:

[0012] The constraint setting module is used to acquire the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths. The environmental BIM model includes the main structure BIM data and the piping BIM data.

[0013] The random design module is used to perform random design of supports and hangers under the constraints of the support and hanger design based on the pipeline BIM data and the main structure BIM data to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers.

[0014] The fitness analysis module is used to perform fitness analysis calculations on the first support bracket design scheme based on the environmental BIM model to obtain the fitness of the first support bracket. The fitness calculation is performed by predicting the first stability and the first material usage of the first support bracket design scheme.

[0015] The optimization module is used to optimize the support and hanger design scheme to obtain the optimal support and hanger scheme as the design result. The optimization is performed by setting an optimization cluster mechanism based on the constraint strength.

[0016] The beneficial effects of this invention are:

[0017] Compared to existing technologies, this invention firstly achieves the automatic and objective extraction of design constraints from BIM models and quantifies their strength, providing precise input for subsequent design. Secondly, by combining random generation with intelligent evaluation, it effectively expands the exploration space of design solutions, helping to discover optimal layouts that transcend traditional experience. Thirdly, it creatively integrates intelligent prediction of the overall stability of supports and hangers with material economic indicators, establishing a scientific and comprehensive scheme evaluation system. Finally, the proposed optimization cluster mechanism can adaptively adjust the optimization search strategy based on constraint strength, improving optimization efficiency and result quality in complex engineering scenarios, ultimately achieving a comprehensive improvement in the safety, economy, and intelligence level of support and hanger design. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a BIM-based support and hanger design method provided by this invention;

[0019] Figure 2 This is a structural schematic diagram of a BIM-based support and hanger design system provided by the present invention.

[0020] In the attached diagram, the components represented by each number are as follows:

[0021] Constraint setting module 11, random design module 12, fitness analysis module 13, optimization module 14. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0024] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0025] Example 1, as Figure 1 As shown, this embodiment of the invention provides a BIM-based support and hanger design method, including:

[0026] S10: Obtain the environmental BIM model for support and hanger design, set constraints, and obtain support and hanger design constraints and constraint strengths. The environmental BIM model includes main structure BIM data and piping BIM data.

[0027] Specifically, the environmental BIM model for the support and hanger design is obtained, constraints are set, and the support and hanger design constraints and constraint strengths are obtained, including:

[0028] Obtain the environmental BIM model for support and hanger design, which includes the main structure BIM data and the piping BIM data;

[0029] Based on the main structure BIM data and pipeline BIM data, the pipeline support and hanger requirements are identified to obtain the required number of supports and hangers, which serves as a design constraint for the supports and hangers.

[0030] Based on the support and hanger design data of the same pipeline family, the constraint strength of the support and hanger design constraints is analyzed.

[0031] First, acquire the digital engineering model required for support and hanger design, namely the environmental BIM model. This environmental BIM model is an integrated three-dimensional information carrier containing two types of core data: one is the main structural BIM data reflecting the geometry and attribute information of load-bearing components such as beams, slabs, and columns of the building; the other is the piping BIM data reflecting the spatial layout and physical characteristics of various pipes, air ducts, and other electromechanical pipelines to be supported. Acquiring the main structural BIM data and piping BIM data is the foundation for all subsequent analysis and design work.

[0032] Secondly, based on the BIM data of the main structure and the piping, the requirements for pipe support hangers and supports are identified to determine the required number of hangers and supports, which serves as a design constraint. This identification process analyzes factors such as the pipe's direction, span, load, and available anchorage locations in the main structure to calculate the minimum number of hangers and supports required to ensure the safe and stable operation of the pipeline system. The calculated required number of hangers and supports will serve as the core constraint for subsequent hanger and support layout design.

[0033] Specifically, based on the main structure BIM data and the piping BIM data, the demand for pipe support hangers is identified to obtain the required number of hangers, including:

[0034] Based on historical support and hanger design data, collect sample main structure BIM data set and sample piping BIM data set, and collect the number of supports and hangers designed under different sample main structure BIM data and sample piping BIM data, and mark to obtain sample support and hanger quantity set;

[0035] Construct a deep learning-based support and hanger demand predictor, where the input features are the main structure BIM data and the piping BIM data, and the output feature is the required number of supports and hangers;

[0036] The support and hanger demand predictor is trained under supervision using the sample main structure BIM data set, the sample pipe BIM data set, and the sample support and hanger quantity set.

[0037] Input the main structure BIM data and piping BIM data into the support and hanger demand predictor, and output the required number of supports and hangers.

[0038] First, two key information samples were systematically collected from historical support and hanger design data. The historical support and hanger design data consists of complete design documents for previous projects that have been designed, constructed, or verified by engineering projects. These documents include the project's 3D building information model, design drawings, calculation sheets, and final support and hanger layout as-built documentation. This data was obtained through standardized extraction and organization of existing project archive databases.

[0039] Specifically, two key information samples are collected from historical support and hanger design data. The first part is a set of sample main structure BIM data reflecting the engineering structural conditions. The second part is a set of sample piping BIM data corresponding to the structure. Simultaneously, for each set of corresponding sample main structure BIM data and sample piping BIM data, the actual or verified number of supports and hangers used in the original design scheme is precisely recorded, and this number is used as a standard for labeling, thus forming a set of sample support and hanger quantities that corresponds one-to-one with the sample data. This set of sample support and hanger quantities serves as the baseline ground truth data for subsequent training, guiding and evaluating the complex mapping relationship between building information model features and reasonable support and hanger quantities, ensuring the engineering rationality and reliability of the output results.

[0040] Secondly, a support and hanger demand predictor is constructed based on a deep learning framework. The input features of this predictor are the main structure BIM data and piping BIM data to be analyzed, and the output feature is the required number of supports and hangers. Specifically, the constructed support and hanger demand predictor is trained under supervised conditions using the aforementioned sample main structure BIM data set, sample piping BIM data set, and sample support and hanger quantity set.

[0041] The training process continuously adjusts the model's internal parameters to minimize the difference between the number of supports and hangers predicted by the model and the actual number of supports and hangers labeled. This allows the model to gradually master the ability to accurately infer the reasonable number of supports and hangers based on the given building information model conditions. The training continues until the model's performance on the validation set reaches the preset standard and converges, resulting in a support and hanger demand predictor with practical engineering accuracy and good generalization ability.

[0042] The convergence condition is primarily set based on the stability of the model's prediction error on independent validation datasets. For example, if the mean absolute error on the validation set remains at or below the order of two supports for 15 consecutive training cycles, and the prediction accuracy reaches 95% or higher, the model is considered converged and training is terminated. The core advantage of this support demand predictor lies in its ability to automatically learn the inherent nonlinear mapping relationship between the complex 3D space and attribute information composed of main structure BIM data and pipeline BIM data and the specific number of supports. This replaces traditional estimation methods that rely on manual experience and simplified formulas, providing an objective, accurate, and reproducible benchmark for the number of supports under various complex pipeline layouts and structural conditions, thus improving the accuracy and reliability of design inputs.

[0043] For example, since there is a highly nonlinear and complex correlation between the geometric topology, spatial relationships, and engineering attributes of the main structure BIM data and the piping BIM data and the required number of supports and hangers, and deep learning models have advantages in high-dimensional feature extraction and complex pattern fitting, a deep learning model can be selected to construct the support and hanger demand predictor.

[0044] Specifically, this support and hanger demand predictor mainly consists of an input layer, a fusion feature extraction layer, and a regression output layer. The input layer receives two sets of input data in parallel: one set is normalized BIM data of the main structure, including the geometric dimensions, material properties, and spatial coordinate matrix of structural components; the other set is normalized BIM data of the pipelines, including the pipeline's direction, diameter, material, medium, and spatial arrangement vector. The fusion feature extraction layer adopts a dual-branch deep neural network structure, with each branch processing one type of BIM data. Each branch contains three fully connected layers, with the number of hidden layer neurons adaptively configured according to the dimension of the input features. Each neural network layer uses the ReLU activation function to introduce non-linear transformation capability, and Dropout layers are embedded between network layers with a dropout rate set to 0.3 to effectively suppress model overfitting and improve its generalization performance. The features extracted by the dual branches are concatenated at a specific level and then fused and abstracted through two fully connected layers. The output layer uses a linear activation function to map the final abstract features to a continuous integer value, which serves as the predicted demand for supports and hangers.

[0045] During training, key hyperparameters included a learning rate of 0.0005, 200 training epochs, and a batch size of 32. The learning rate was set to balance training stability and convergence speed; the number of training epochs ensured the model fully learned the mapping relationship between BIM data and the number of supports and hangers; and the batch size balanced training efficiency with memory resource consumption. Specifically, a supervised learning training method was adopted. Sample main structure BIM data sets and sample pipeline BIM data sets were collected from historical support and hanger design data as the input sample set. Simultaneously, the corresponding sample support and hanger quantity sets were obtained to form a sample label set. The input sample set and the corresponding label sample set were divided into training, validation, and test sets in a 7:2:1 ratio.

[0046] Furthermore, the BIM data of the main structure and the pipeline system are used as parallel inputs, with the corresponding number of supports and hangers as the supervision signal. The network weight parameters are iteratively optimized using a backpropagation algorithm and an Adam optimizer. The mean squared error loss function is used to measure the deviation between the predicted and actual number of supports and hangers, and the training process is monitored using a validation set. Training is terminated when the validation set loss function value no longer decreases after 20 consecutive training epochs, and the average absolute error of the model's prediction is less than or equal to 2 supports and hangers, resulting in a converged support and hanger demand predictor. This support and hanger demand predictor can effectively capture the complex nonlinear influence of the multidimensional characteristics of the main structure and pipeline system on support demand, achieving accurate prediction of the number of supports and hangers.

[0047] It is worth noting that, in another optional embodiment, the input features of the support and hanger demand predictor can be a BIM model image. Specifically, a 3D BIM model containing the current piping system and main structure is rendered into a BIM model image according to a preset global perspective and display rules, such as displaying pipe lines, structural outlines, and hiding non-critical components. This BIM model image comprehensively represents the spatial layout, routing density, and structural environment of the piping. Accordingly, the support and hanger demand predictor can be constructed as an image regression model based on a convolutional neural network. Through training, it learns to directly predict the required number of supports and hangers from the features of the BIM model image.

[0048] Finally, the trained support and hanger demand predictor is used for actual prediction. The current main structure BIM data and piping BIM data to be analyzed are input into the trained support and hanger demand predictor, which calculates and outputs the required number of supports and hangers. This required number of supports and hangers serves as the support and hanger design constraint.

[0049] Furthermore, to assess the stringency or adjustability of the aforementioned support and hanger design constraints, it is necessary to analyze their constraint strength. Specifically, constraint strength represents the degree of deviation between current design requirements and historical engineering experience. Higher constraint strength indicates a high degree of agreement between current requirements and historical experience, indicating stricter design constraints. During optimization, the space for adjusting the position and parameters of supports and hangers while maintaining the required number of supports and hangers remains relatively limited. Conversely, lower constraint strength indicates a significant difference between current requirements and historical experience, indicating relatively looser design constraints and greater potential for layout optimization.

[0050] Specifically, based on the support and hanger design data of the same pipeline family, the constraint strength of the support and hanger design constraints is analyzed, including:

[0051] Analyze the similarity between the pipeline BIM data and historical pipeline BIM data within a historical period, and select the historical pipeline BIM data with the highest similarity as multiple pipeline BIM data of the same family;

[0052] Obtain the number of supports and hangers installed for multiple pipe families in BIM data;

[0053] The ratio of the required number of supports and hangers to the average number of supports and hangers installed in multiple families is used as the constraint strength.

[0054] First, the similarity between the current project's piping BIM data and historical piping BIM data from previous periods is analyzed, and data is filtered based on the similarity score. Specifically, historical piping BIM data from previous periods is obtained through the systematic collection, organization, and archiving of multiple completed Building Information Modeling (BIM) project files whose design results have been validated in engineering practice. Each historical piping BIM data set contains the geometric information, spatial coordinates, physical attributes, and topological connections of all components constituting the piping system, representing a precise three-dimensional digital model of the piping network and its engineering status in the corresponding historical engineering project.

[0055] Secondly, the analysis process involves calculating the degree of feature matching between the current pipeline BIM data and historical pipeline BIM data across multiple dimensions, including geometric dimensions, spatial orientation, connectivity, and engineering attributes. Similarity calculation typically employs feature extraction and comparison methods. For example, the coordinates of key nodes in the pipeline network, pipe segment direction vectors, pipe diameter sequences, and material types are parameterized into a set of feature vectors. A quantified similarity score is obtained by calculating the Euclidean distance or cosine similarity between the current feature vector and each historical feature vector.

[0056] Then, based on the calculated similarity scores, the historical pipeline BIM data with the highest similarity scores to the current pipeline BIM data are selected as multiple family pipeline BIM data for the current pipeline system. This screening step aims to locate the reference cases most similar to the current design scenario from historical experience, providing a directly relevant comparison benchmark for subsequent constraint strength quantification.

[0057] Secondly, obtain the installation quantity of supports and hangers corresponding to each of the above-mentioned multiple family pipeline BIM data. For each family pipeline BIM data, extract the actual installation quantity of supports and hangers used in the pipeline system from the corresponding historical design records, and collect the installation quantities of multiple family pipeline supports and hangers.

[0058] Finally, the ratio of the required number of supports and hangers to the average number of supports and hangers installed within the same family is calculated, and this ratio is defined as the constraint strength. The specific calculation process involves first taking the arithmetic mean of the obtained installation quantities of multiple supports and hangers within the same family; this average represents the typical support and hanger usage level in similar historical cases. Then, the required number of supports and hangers obtained through the support and hanger demand forecaster is compared with this historical average, and the ratio between the two is calculated. The resulting value is the constraint strength. This constraint strength directly characterizes the degree of agreement between current design requirements and historical engineering experience.

[0059] S20: Based on the pipeline BIM data and the main structure BIM data, random design of the supports and hangers is carried out under the support and hanger design constraints to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers.

[0060] Specifically, based on the pipeline BIM data and the main structure BIM data, a random design of the supports and hangers is performed under the constraints of the support and hanger design to obtain a first support and hanger design scheme, including:

[0061] Based on the pipeline BIM data and the main structure BIM data, the spatial location of the hanging support is obtained;

[0062] According to the design constraints of the support and hanger, the design position and design parameters of the hanger are randomly set within the space of the hanger position to obtain multiple first design positions and multiple first design parameters, which serve as the first support and hanger design scheme.

[0063] First, the location space of the hangers needs to be obtained. Specifically, based on the pipe BIM data, the routing and distribution of the pipes in three-dimensional space are analyzed to identify the continuous pipe segments that need support. Simultaneously, the main structure BIM data is analyzed to identify load-bearing components suitable for installing hangers, such as the surfaces and surrounding space of beams, slabs, columns, or shear walls. The final obtained hanger location space is defined as the set of three-dimensional spaces below or to the side of the pipe segment area that needs support, and which have the potential for effective connection with the load-bearing components of the main structure. This hanger location space is a potential installation range that can be continuous or composed of multiple discrete areas.

[0064] Secondly, within the acquired space of support locations, the required number of supports is randomly arranged and configured according to the design constraints. For each support to be arranged, the following operations are performed: Within the support location space, a three-dimensional coordinate point is generated using a random algorithm as the first design position for that support. This three-dimensional coordinate point must ensure that the support can effectively connect the pipe above to the main structure below or to the side. Simultaneously, a set of design parameters is randomly selected from a predefined design parameter library for this support as its first design parameter. This predefined design parameter library is a database systematically constructed by integrating industry standards and specifications, manufacturer product catalogs, and commonly used selections from historical projects. It includes various standard support models and their complete technical parameters applicable to different pipe diameters, different medium loads, different seismic resistance levels, and different installation scenarios. The first design parameter typically includes the support type, model, material, cross-sectional dimensions, length of the main components, and specifications of the connectors. The selection of the support type and model must match the pipe characteristics and load requirements.

[0065] Through the above process, random design locations and parameters can be generated for each required support / hanger. The set of first design locations and first design parameters for all supports / hangers constitutes a complete first support / hanger design scheme. This first support / hanger design scheme is an initial design solution generated through a random process under the conditions of satisfying spatial feasibility and basic type matching, and can serve as the search starting point for subsequent fitness evaluation and optimization processes.

[0066] S30: Based on the environmental BIM model, perform fitness analysis calculations on the first hanger design scheme to obtain the fitness of the first hanger, wherein the fitness calculations include predicting the first stability and the first material usage of the first hanger design scheme.

[0067] Specifically, based on the environmental BIM model, an adaptability analysis calculation is performed on the design scheme of the first hanger to obtain the adaptability of the first hanger, including:

[0068] Based on the multiple first design locations, multiple first location structural BIM data and multiple first location pipeline BIM data are obtained by indexing within the environmental BIM model;

[0069] Multiple structural BIM data and multiple pipeline BIM data at the first location are combined with multiple first design parameters, input into the stability analyzer, and multiple first location stability values ​​are obtained from the output. The average value is then calculated to obtain the first stability value.

[0070] Based on multiple primary design parameters and the required number of supports and hangers within the design constraints, the primary material consumption is calculated.

[0071] Based on the first material usage amount and the preset material usage amount, the first material saving degree is calculated, and based on the first stability degree, the first hanger adaptability degree is calculated.

[0072] First, based on the multiple first design locations included in the first support bracket design scheme, data indexing is performed in the environmental BIM model. For each first design location, BIM data fragments of the main structure within the surrounding space of that first design location are extracted from the environmental BIM model, serving as the first location structural BIM data corresponding to that first design location. Simultaneously, BIM data fragments of the associated piping sections that need to be supported are extracted, serving as the first location piping BIM data corresponding to that first design location.

[0073] Secondly, the multiple sets of structural BIM data and piping BIM data obtained above for the first location are combined with the corresponding first design parameters of the supports and hangers, and input into a pre-trained stability analyzer. This stability analyzer is a computational model built on a deep neural network, mapping the complex nonlinear relationship between the specific local structural environment, piping load conditions, and support and hanger design parameters and the safety and stability performance of the corresponding support and hanger system. Based on the input local environment data and support and hanger parameters, the stability analyzer can calculate and output the stability score of each support and hanger under the specific working condition, i.e., the first location stability. The arithmetic mean of the first location stability of all supports and hangers is calculated to obtain the first stability, which characterizes the overall safety performance of the entire first support and hanger design scheme. Specifically, the higher the first stability value, the better the overall performance of the first support and hanger design scheme in terms of structural safety.

[0074] Specifically, the training steps for the stability analyzer include:

[0075] Based on the installation and test data of the supports and hangers, the following sets of BIM data were collected: structural BIM data set, piping BIM data set, design parameter set, and stability set obtained from the test.

[0076] Construct a stability analyzer based on deep learning, where the input features are location structure BIM data, location pipeline BIM data and design data, and the output feature is stability.

[0077] The stability analyzer is supervised and tested using the sample location structure BIM data set, sample location pipeline BIM data set, sample design parameter set, and sample stability set until the test accuracy is qualified, thus completing the training.

[0078] First, based on the installation test data of the supports and hangers, datasets for training and testing are collected. The installation test data originates from actual mechanical performance test records of support and hanger systems in completed projects, or design cases verified through rigorous finite element analysis. From the installation test data, for each test sample point, three parts of information are extracted: first, a fragment of structural BIM data around the installation location of the support and hanger, serving as the structural BIM data for the sample location; second, a fragment of BIM data of the pipe section supported by the support and hanger, serving as the pipe BIM data for the sample location; and third, the design parameters of the support and hanger, such as its type, size, and materials, serving as the sample design parameters. Simultaneously, indicators quantifying the safety performance of the test sample point, obtained through actual measurement or simulation calculations, such as safety factors or displacement limit compliance, are recorded as the true stability label for the sample. By systematically collecting multiple sample points, a set of structural BIM data for the sample location, a set of pipe BIM data for the sample location, a set of sample design parameters, and a corresponding set of sample stability can be constructed.

[0079] Secondly, a stability analyzer based on deep learning is constructed. The input layer of this stability analyzer is designed to receive three sets of parallel feature data, including location structure BIM data, location piping BIM data, and support and hanger design parameter data. The output layer is designed to generate a continuous value, namely the stability.

[0080] Furthermore, the aforementioned sample location structure BIM data set, sample location pipeline BIM data set, and sample design parameter set are used as input features, and the sample stability set is used as supervision labels. The stability analyzer is subjected to supervised training and testing until the test accuracy is satisfactory, thus completing the training. Specifically, the satisfactory test accuracy criteria are set based on the model's predictive performance indicators on independent test sets. For example, the average absolute percentage error between the model's predicted stability value and the true stability label is required to be less than 5%, and the Pearson correlation coefficient between the predicted value and the true value is required to be higher than 0.95, at which point the model training is considered converged and the accuracy is satisfactory. The trained stability analyzer is obtained. This stability analyzer can quickly and automatically predict the quantitative stability of the corresponding support system based on the local structural environment, pipeline load conditions, and support design parameters. It effectively replaces the traditional evaluation method that relies on complex manual calculations or finite element simulations. It is used to efficiently and batch evaluate the mechanical safety performance of massive candidate support design schemes during the scheme optimization iteration process, thereby providing key safety input indicators for fitness calculations.

[0081] For example, since there is a highly nonlinear and complex correlation between the system stability, which is jointly determined by the local structural environment, pipeline load, and support design parameters, and the aforementioned input features, and since deep learning models have advantages in extracting deep features and establishing complex mappings from high-dimensional heterogeneous data, a deep learning-based model is chosen to construct this stability analyzer, such as convolutional neural networks, graph neural networks, and deep fully connected networks. A detailed explanation will be given using a deep fully connected network as an example.

[0082] Specifically, this stability analyzer mainly consists of an input layer, a feature fusion abstraction layer, and a regression output layer. The input layer receives three sets of normalized input vectors in parallel: the first set is a feature vector extracted from the sample location's structural BIM data, containing information such as structural component geometry and material properties; the second set is a feature vector extracted from the sample location's pipeline BIM data, containing information such as pipeline geometry and medium properties; and the third set is a sample design parameter vector, containing parameters such as support type, size, and material. The feature fusion abstraction layer employs a multi-layer fully connected neural network structure. It first concatenates the three sets of input vectors, then performs nonlinear transformation and feature fusion through multiple fully connected layers. The number of hidden layer neurons is adaptively configured based on the total feature dimension after concatenation, for example, by decreasing layer by layer. Each neural network layer uses the ReLU activation function to introduce nonlinear transformation capability, and Dropout layers are embedded between network layers with a dropout rate set to 0.3 to effectively suppress model overfitting and improve its generalization performance. The output layer uses a linear activation function to map the final abstracted high-level features to a continuous numerical value as the predicted stability.

[0083] During training, key hyperparameters included a learning rate of 0.0001, 300 training epochs, and a batch size of 16. The learning rate was set to balance training stability and convergence speed; the number of training epochs ensured the model fully learned the complex relationship between input features and stability; and the batch size balanced training accuracy with memory resource consumption. Specifically, supervised learning was employed, using the collected sample location structure BIM data set, sample location pipeline BIM data set, and sample design parameter set as the input sample set, and the sample stability set as the supervision label set. The input sample set and the corresponding label sample set were divided into training, validation, and test sets in a 7:2:1 ratio.

[0084] Furthermore, the feature vectors of the samples in the training set are used as input, and the corresponding stability labels are used as supervision signals. The network weight parameters are iteratively optimized using the backpropagation algorithm and the Adam optimizer. The mean squared error loss function is used to measure the deviation between the predicted stability and the actual stability label, and the training process is monitored using a validation set. Training is terminated when the validation set loss function value no longer decreases for 30 consecutive training epochs, the mean absolute percentage error between the model's predicted stability and the true stability on the test set is less than 5%, and the coefficient of determination R² is greater than 0.95, resulting in a converged stability analyzer.

[0085] It is important to note that, in another optional embodiment, the input features of the stability analyzer can be a BIM model image of the design location area. Specifically, for each support / hanger design location to be evaluated, a three-dimensional spatial range is extracted centered on that location. The main structure BIM data and piping BIM data within this three-dimensional spatial range are then rendered to generate a BIM model image of the design location area. This BIM model image of the design location area clearly shows the specific structural conditions, piping connections, and spatial spacing at the proposed installation point of the support / hanger. Accordingly, the stability analyzer can be constructed as an image regression model based on a convolutional neural network. Through training, it learns from the fused features of the BIM model image of the design location area and the support / hanger design parameters to predict the stability of the support / hanger at that location.

[0086] Finally, multiple first-location structural BIM data and multiple first-location pipeline BIM data are combined with multiple first design parameters and input into the trained stability analyzer. The output obtains multiple first-location stability values, and the average value is calculated to obtain the first stability value.

[0087] Furthermore, based on all the first design parameters in the first support / hanger design scheme, the material consumption of each support / hanger, such as the volume or weight of steel, can be calculated. By summing the individual material consumption of all supports / hangers and combining it with the required number of supports / hangers already determined in the support / hanger design constraints, the total first material consumption required to support the entire piping system can be calculated. To evaluate the economic efficiency of this total first material consumption, it is compared with a preset material consumption. This preset material consumption is a material consumption reference benchmark pre-set based on engineering experience and project requirements, representing the acceptable or expected material consumption level for the project while meeting basic safety and functional requirements. It is based on consumption estimated using empirical formulas, average consumption of similar historical projects, or design limits.

[0088] The first material saving degree is obtained through comparative calculation. Specifically, the first material saving degree = (preset material usage - total first material usage) / preset material usage, which reflects the degree of material saving of the current design scheme relative to the preset benchmark. The larger the value of the first material saving degree, the more economical the material use.

[0089] Finally, the first stability index, representing safety, and the first material saving index, representing economy, are combined using a weighted calculation formula to obtain the first hanger adaptability. Specifically, the first hanger adaptability = α × first stability index + β × first material saving index. Here, α and β are weighting coefficients, set according to the priority requirements of safety and economy in the engineering design objectives. For example, in a scenario prioritizing safety, α = 0.7 and β = 0.3 can be set; in a scenario emphasizing cost control, α = 0.4 and β = 0.6 can be set. The sum of the weighting coefficients is 1. The calculated first hanger adaptability is used to characterize the comprehensive performance level of the design scheme in terms of the two core objectives of safety and economy. The higher the first hanger adaptability score, the more economical and reasonable the material use and the better the overall performance of the first hanger design scheme while meeting structural safety requirements.

[0090] S40: Optimize the support and hanger design scheme to obtain the optimal support and hanger scheme as the design result. In this process, an optimization cluster mechanism is set up based on the constraint strength to perform optimization.

[0091] Specifically, the design scheme of the supports and hangers is optimized to obtain the optimal support and hanger scheme, which is the design result, including:

[0092] Continue to randomly design support and hanger schemes, analyze and calculate the adaptability of support and hangers, until the first preset number of optimizations is reached, and obtain multiple support and hanger design schemes;

[0093] Multiple support and hanger design schemes are clustered to obtain multiple support and hanger design scheme clusters, and the mean support and hanger fitness of multiple support and hanger design scheme clusters is calculated. Clustering is performed based on the coordinate similarity of multiple design positions within each support and hanger design scheme.

[0094] Select the support and hanger design scheme cluster with the largest average fitness value as the selected support and hanger design scheme cluster;

[0095] Based on the constraint strength, the selected support and hanger design scheme family is expanded to obtain the design optimization range;

[0096] Continue iterative optimization within the design optimization range until the second preset number of optimizations is reached, the optimization converges, and the optimal support and hanger scheme with the greatest adaptability is output as the design result.

[0097] First, continuing the initial random design process, a large number of new support and hanger design schemes are generated. For each newly generated support and hanger design scheme, its fitness is calculated according to the aforementioned method. This process continues until the total number of generated schemes reaches a first preset number of optimizations. This first preset number of optimizations is set based on the balance between the complexity of the problem to be optimized and computational resources. For example, for a medium-complexity piping system, the first preset number of optimizations can be set to 5000 to ensure sufficient exploration in a broad solution space and obtain a sufficiently diverse set of candidate design schemes.

[0098] Secondly, cluster analysis is performed on the generated multiple support and hanger design schemes. Specifically, the clustering operation is based on the set of three-dimensional coordinates of all support and hanger design locations in each scheme. By calculating the similarity of coordinate distributions between different support and hanger design schemes, such as based on the centroid distance of the set of all support and hanger location points or using a density-based clustering algorithm, the multiple support and hanger design schemes are divided into multiple distinct clusters, each called a support and hanger design scheme cluster. The support and hanger design schemes within each cluster exhibit high similarity in the spatial layout of the supports and hangers. Furthermore, the arithmetic mean of the support and hanger fitness of all support and hanger design schemes within each cluster is calculated to obtain the mean support and hanger fitness of that cluster. This mean support and hanger fitness is used to characterize the average comprehensive performance level of the support and hanger design schemes corresponding to a specific spatial layout pattern in terms of safety and economy.

[0099] From all support and hanger design schemes, the scheme with the highest average fitness is selected and designated as the current optimal region representative, i.e., the selected support and hanger design scheme cluster. Then, based on the constraint strength obtained from the aforementioned calculations, this selected support and hanger design scheme cluster is expanded to obtain the design optimization range. This design optimization range refers to the set of allowed design position coordinates for each support and hanger to be deployed during the next stage of fine-tuning iterations. It is used to constrain and guide subsequent scheme generation, ensuring the search focuses on the neighborhood of the most promising solution space.

[0100] Specifically, based on the constraint strength, the selected support and hanger design scheme family is expanded to obtain a design optimization range, including:

[0101] Obtain the position coordinate range of the required number of supports within the selected support and hanger design scheme cluster, as multiple support and hanger ranges, where each support and hanger range includes the position coordinates of two endpoints;

[0102] Using the aforementioned constraint strength, the range of multiple supports and hangers is adjusted to obtain the design optimization range, wherein the adjustment includes enlarging or reducing the range of supports and hangers.

[0103] First, obtain the position coordinate range of each support within the selected support design scheme cluster. For each support in the required number of supports specified by the support design constraints, analyze its design position within all support design schemes in the selected support design scheme cluster. Extract the coordinate values ​​of the design position point in the primary consideration direction. Since supports are mainly arranged along the axial direction of the pipeline, the primary consideration direction is set as the axial direction of the pipeline. Determine the minimum and maximum coordinate values ​​from the extracted coordinate values; these minimum and maximum coordinate values ​​constitute the two endpoint position coordinates of the support within the selected support design scheme cluster. Perform the above operation for each support in the required number of supports to obtain multiple support ranges equal to the required number of supports.

[0104] Secondly, the sizing of each support / hanger is adjusted using the constraint strength. Specifically, the adjustment method is linear scaling, calculating the original length of each support / hanger, which is the difference between the coordinates of its two endpoints. This original length is multiplied by the constraint strength coefficient to obtain the adjusted new length. Then, using the center point of the original length as a reference, the new length is symmetrically distributed to both ends, and the coordinates of the adjusted new endpoints are calculated.

[0105] Exemplarily, assume that the number of required support hangers is 10. For the first support hanger, the minimum axial coordinate among all the solutions in the selected support hanger design solution cluster is x1, and the maximum is x2, and x1 < x2. Let the original range length L_original = x2 - x1 = 10 units. If the calculated constraint strength is 1.1, then the adjusted new range length L_new = L_original × 1.1 = 11 units. The coordinate of the center point of the original range is (x1 + x 2） / 2. Then the adjusted new endpoint coordinates are: x1_new = center point - L_new / 2, x2_new = center point + L_new / 2. It can be calculated that the new range extends 0.5 units towards both ends compared to the original range.

[0106] Specifically, when the constraint strength is greater than 1, the range of the support hanger is enlarged, allowing exploration in a wider neighborhood, which is applicable to scenarios with relatively low constraint strength and a large optimization space; when the constraint strength is less than 1, the range of the support hanger is reduced, making the search more concentrated on the core area where high-fitness solutions gather, which is applicable to scenarios with relatively high constraint strength and strict design limitations. The jointly adjusted new coordinate ranges of all support hangers constitute the design optimization range for the next stage, that is, each support hanger generates its position within the interval defined by its corresponding new endpoint coordinates in the design optimization range. In summary, this method realizes dynamically guiding the focus degree of the optimization search according to the strictness of the design constraints.

[0107] Furthermore, within the determined design optimization range, the optimization iteration is started. Among them, the iterative optimization process can be implemented by using swarm intelligence optimization algorithms, such as genetic algorithms and particle swarm algorithms.

[0108] Taking a genetic algorithm as an example, the process specifically includes: First, the design optimization range is encoded into a solution space, and a population consisting of multiple support and hanger design schemes is randomly initialized within it, with each design scheme encoded as an individual. In each generation iteration, evaluation, selection, crossover and mutation, and update operations are performed. The evaluation operation calculates the support and hanger fitness of each individual in the population. The selection operation selects superior individuals as parents based on their fitness. The crossover and mutation operation crossovers the selected parent individuals to combine the advantages of different schemes and mutates the individual genes to generate new explorations within the solution space. The update operation uses the newly generated offspring individuals to update the current population, forming a new generation. This iterative process continues, driving the performance of the entire population's schemes towards higher fitness, until the number of iterations reaches a second preset optimization number or the convergence condition such as the population fitness stabilizing is met. The second preset optimization number is set based on the computational load required for refined searching and the optimization accuracy requirements of actual engineering. For example, the second preset optimization number can be set to 2000 to ensure sufficient searching within the focused high-potential solution space while balancing computational time. Finally, from all the support and hanger design schemes evaluated throughout the optimization process, the single scheme with the highest support and hanger adaptability score is selected as the optimal support and hanger scheme output, which becomes the final design result.

[0109] In summary, the embodiments of this application have at least the following technical effects:

[0110] Compared to existing technologies, this application firstly achieves intelligent and data-driven design processes. Through deep learning models, it automatically extracts design constraints and evaluates stability from BIM data, reducing reliance on experience and improving the objectivity and accuracy of the design. Secondly, it achieves multi-objective collaborative optimization of safety and economy. By defining a comprehensive fitness index and a two-stage optimization strategy, it effectively reduces material costs while ensuring structural safety. Thirdly, by dynamically adjusting the search range through constraint strength, it improves optimization efficiency, and the output solution has a certain tolerance for construction errors, enhancing the adaptability and engineering robustness of the optimization algorithm. Finally, this application constructs a complete automated design process, from BIM input to solution output, improving the design efficiency of supports and hangers for complex pipeline systems.

[0111] Example 2, as Figure 2 As shown, based on the same inventive concept as the BIM-based support and hanger design method provided in Embodiment 1, this embodiment of the invention also provides a BIM-based support and hanger design system, including:

[0112] The constraint setting module 11 is used to acquire the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths. The environmental BIM model includes the main structure BIM data and the piping BIM data.

[0113] The random design module 12 is used to perform random design of supports and hangers under the constraints of the support and hanger design based on the pipeline BIM data and the main structure BIM data to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers.

[0114] The fitness analysis module 13 is used to perform fitness analysis calculations on the first support bracket design scheme based on the environmental BIM model to obtain the fitness of the first support bracket, wherein the fitness calculation is performed by predicting the first stability and the first material usage of the first support bracket design scheme.

[0115] The optimization module 14 is used to optimize the support and hanger design scheme to obtain the optimal support and hanger scheme as the design result. The optimization is performed by setting an optimization cluster mechanism according to the constraint strength.

[0116] Specifically, the constraint setting module 11 is used for:

[0117] Obtain the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths, including:

[0118] Obtain the environmental BIM model for support and hanger design, which includes the main structure BIM data and the piping BIM data;

[0119] Based on the main structure BIM data and pipeline BIM data, the pipeline support and hanger requirements are identified to obtain the required number of supports and hangers, which serves as a design constraint for the supports and hangers.

[0120] Based on the support and hanger design data of the same pipeline family, the constraint strength of the support and hanger design constraints is analyzed.

[0121] Specifically, based on the main structure BIM data and the piping BIM data, the demand for pipe support hangers is identified to obtain the required number of hangers, including:

[0122] Based on historical support and hanger design data, collect sample main structure BIM data set and sample piping BIM data set, and collect the number of supports and hangers designed under different sample main structure BIM data and sample piping BIM data, and mark to obtain sample support and hanger quantity set;

[0123] Construct a deep learning-based support and hanger demand predictor, where the input features are the main structure BIM data and the piping BIM data, and the output feature is the required number of supports and hangers;

[0124] The support and hanger demand predictor is trained under supervision using the sample main structure BIM data set, the sample pipe BIM data set, and the sample support and hanger quantity set.

[0125] Input the main structure BIM data and piping BIM data into the support and hanger demand predictor, and output the required number of supports and hangers.

[0126] Specifically, based on the support and hanger design data of the same pipeline family, the constraint strength of the support and hanger design constraints is analyzed, including:

[0127] Analyze the similarity between the pipeline BIM data and historical pipeline BIM data within a historical period, and select the historical pipeline BIM data with the highest similarity as multiple pipeline BIM data of the same family;

[0128] Obtain the number of supports and hangers installed for multiple pipe families in BIM data;

[0129] The ratio of the required number of supports and hangers to the average number of supports and hangers installed in multiple families is used as the constraint strength.

[0130] The random design module 12 is specifically used for:

[0131] Based on the pipeline BIM data and the main structure BIM data, a random design of the supports and hangers is performed under the constraints of the support and hanger design to obtain a first support and hanger design scheme, including:

[0132] Based on the pipeline BIM data and the main structure BIM data, the spatial location of the hanging support is obtained;

[0133] According to the design constraints of the support and hanger, the design position and design parameters of the hanger are randomly set within the space of the hanger position to obtain multiple first design positions and multiple first design parameters, which serve as the first support and hanger design scheme.

[0134] The fitness analysis module 13 is specifically used for:

[0135] Based on the environmental BIM model, a fitness analysis calculation is performed on the design scheme of the first hanger to obtain the fitness of the first hanger, including:

[0136] Based on the multiple first design locations, multiple first location structural BIM data and multiple first location pipeline BIM data are obtained by indexing within the environmental BIM model;

[0137] Multiple structural BIM data and multiple pipeline BIM data at the first location are combined with multiple first design parameters, input into the stability analyzer, and multiple first location stability values ​​are obtained from the output. The average value is then calculated to obtain the first stability value.

[0138] Based on multiple primary design parameters and the required number of supports and hangers within the design constraints, the primary material consumption is calculated.

[0139] Based on the first material usage amount and the preset material usage amount, the first material saving degree is calculated, and based on the first stability degree, the first hanger adaptability degree is calculated.

[0140] Specifically, the training steps for the stability analyzer include:

[0141] Based on the installation and test data of the supports and hangers, the following sets of BIM data were collected: structural BIM data set, piping BIM data set, design parameter set, and stability set obtained from the test.

[0142] Construct a stability analyzer based on deep learning, where the input features are location structure BIM data, location pipeline BIM data and design data, and the output feature is stability.

[0143] The stability analyzer is trained and tested using the sample location structure BIM data set, sample location pipeline BIM data set, sample design parameter set, and sample stability set until the test accuracy is qualified, thus completing the training.

[0144] The optimization module 14 is specifically used for:

[0145] Specifically, the design scheme of the supports and hangers is optimized to obtain the optimal support and hanger scheme, which is the design result, including:

[0146] Continue to randomly design support and hanger schemes, analyze and calculate the adaptability of support and hangers, until the first preset number of optimizations is reached, and obtain multiple support and hanger design schemes;

[0147] Multiple support and hanger design schemes are clustered to obtain multiple support and hanger design scheme clusters, and the mean support and hanger fitness of multiple support and hanger design scheme clusters is calculated. Clustering is performed based on the coordinate similarity of multiple design positions within each support and hanger design scheme.

[0148] Select the support and hanger design scheme cluster with the largest average fitness value as the selected support and hanger design scheme cluster;

[0149] Based on the aforementioned constraint strength, the selected support and hanger design scheme family is expanded to obtain the design optimization range;

[0150] Continue iterative optimization within the design optimization range until the second preset number of optimizations is reached, the optimization converges, and the optimal support and hanger scheme with the greatest adaptability is output as the design result.

[0151] Specifically, based on the constraint strength, the selected support and hanger design scheme family is expanded to obtain a design optimization range, including:

[0152] Obtain the position coordinate range of the required number of supports within the selected support and hanger design scheme cluster, as multiple support and hanger ranges, where each support and hanger range includes the position coordinates of two endpoints;

[0153] Using the aforementioned constraint strength, the range of multiple supports and hangers is adjusted to obtain the design optimization range, wherein the adjustment includes enlarging or reducing the range of supports and hangers.

[0154] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0155] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0156] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A BIM-based support and hanger design method, characterized in that, The method includes: Obtain the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths. The environmental BIM model includes the main structure BIM data and the piping BIM data. Based on the pipeline BIM data and the main structure BIM data, the supports and hangers are randomly designed under the constraints of the support and hanger design to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers. Based on the environmental BIM model, the fitness analysis calculation of the first hanger design scheme is performed to obtain the fitness of the first hanger. The fitness calculation includes predicting the first stability and the first material usage of the first hanger design scheme. The design scheme of the support and hanger is optimized to obtain the optimal support and hanger scheme as the design result. In this process, an optimization cluster mechanism is set up based on the constraint strength to perform optimization. Based on the environmental BIM model, a fitness analysis calculation is performed on the design scheme of the first hanger to obtain the fitness of the first hanger, including: Based on the multiple first design locations, multiple first location structural BIM data and multiple first location pipeline BIM data are obtained by indexing within the environmental BIM model; Multiple structural BIM data and multiple pipeline BIM data at the first location are combined with multiple first design parameters, input into the stability analyzer, and multiple first location stability values ​​are obtained from the output. The average value is then calculated to obtain the first stability value. Based on multiple primary design parameters and the required number of supports and hangers within the design constraints, the primary material consumption is calculated. Based on the first material usage amount and the preset material usage amount, the first material saving degree is calculated, and based on the first stability degree, the first hanger adaptability degree is calculated. This includes optimizing the support and hanger design scheme to obtain the optimal support and hanger scheme, which serves as the design result. Continue to randomly design support and hanger schemes, analyze and calculate the adaptability of support and hangers, until the first preset number of optimizations is reached, and obtain multiple support and hanger design schemes; Multiple support and hanger design schemes are clustered to obtain multiple support and hanger design scheme clusters, and the mean support and hanger fitness of multiple support and hanger design scheme clusters is calculated. Clustering is performed based on the coordinate similarity of multiple design positions within each support and hanger design scheme. Select the support and hanger design scheme cluster with the largest average fitness value as the selected support and hanger design scheme cluster; Based on the aforementioned constraint strength, the selected support and hanger design scheme family is expanded to obtain the design optimization range; Continue iterative optimization within the design optimization range until the second preset number of optimizations is reached, the optimization converges, and the optimal support and hanger scheme with the greatest adaptability is output as the design result.

2. The BIM-based support and hanger design method according to claim 1, characterized in that, Obtain the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths, including: Obtain the environmental BIM model for support and hanger design, which includes the main structure BIM data and the piping BIM data; Based on the main structure BIM data and pipeline BIM data, the pipeline support and hanger requirements are identified to obtain the required number of supports and hangers, which serves as a design constraint for the supports and hangers. Based on the support and hanger design data of the same pipeline family, the constraint strength of the support and hanger design constraints is analyzed.

3. The BIM-based support and hanger design method according to claim 2, characterized in that, Based on the main structure BIM data and pipeline BIM data, the pipeline support and hanger requirements are identified to obtain the required number of supports and hangers, including: Based on historical support and hanger design data, collect sample main structure BIM data set and sample piping BIM data set, and collect the number of supports and hangers designed under different sample main structure BIM data and sample piping BIM data, and mark to obtain sample support and hanger quantity set; Construct a deep learning-based support and hanger demand predictor, where the input features are the main structure BIM data and the piping BIM data, and the output feature is the required number of supports and hangers; The support and hanger demand predictor is trained under supervision using the sample main structure BIM data set, the sample pipe BIM data set, and the sample support and hanger quantity set. Input the main structure BIM data and piping BIM data into the support and hanger demand predictor, and output the required number of supports and hangers.

4. The BIM-based support and hanger design method according to claim 2, characterized in that, Based on the support and hanger design data of the same pipeline family, analyze the constraint strength of the support and hanger design constraints, including: Analyze the similarity between the pipeline BIM data and historical pipeline BIM data within a historical period, and select the historical pipeline BIM data with the highest similarity as multiple pipeline BIM data of the same family; Obtain the number of supports and hangers installed for multiple pipe families in BIM data; The ratio of the required number of supports and hangers to the average number of supports and hangers installed in multiple families is used as the constraint strength.

5. The BIM-based support and hanger design method according to claim 1, characterized in that, Based on the pipeline BIM data and the main structure BIM data, a random design of the supports and hangers is performed under the constraints of the support and hanger design to obtain a first support and hanger design scheme, including: Based on the pipeline BIM data and the main structure BIM data, the spatial location of the hanging support is obtained; According to the design constraints of the support and hanger, the design position and design parameters of the hanger are randomly set within the space of the hanger position to obtain multiple first design positions and multiple first design parameters, which serve as the first support and hanger design scheme.

6. The BIM-based support and hanger design method according to claim 5, characterized in that, The training steps for the stability analyzer include: Based on the installation and test data of the supports and hangers, the following sets of BIM data were collected: structural BIM data set, piping BIM data set, design parameter set, and stability set obtained from the test. Construct a stability analyzer based on deep learning, where the input features are location structure BIM data, location pipeline BIM data and design data, and the output feature is stability. The stability analyzer is trained and tested using the sample location structure BIM data set, sample location pipeline BIM data set, sample design parameter set, and sample stability set until the test accuracy is qualified, thus completing the training.

7. The BIM-based support and hanger design method according to claim 2, characterized in that, Based on the aforementioned constraint strength, the selected support and hanger design scheme family is expanded to obtain a design optimization range, including: Obtain the position coordinate range of the required number of supports within the selected support and hanger design scheme cluster, as multiple support and hanger ranges, where each support and hanger range includes the position coordinates of two endpoints; Using the aforementioned constraint strength, the range of multiple supports and hangers is adjusted to obtain the design optimization range, wherein the adjustment includes enlarging or reducing the range of supports and hangers.

8. A BIM-based support and hanger design system, characterized in that, A method for implementing a BIM-based support and hanger design according to any one of claims 1-7 includes: The constraint setting module is used to acquire the environmental BIM model for support and hanger design, set constraints, and obtain the support and hanger design constraints and constraint strengths. The environmental BIM model includes the main structure BIM data and the piping BIM data. The random design module is used to perform random design of supports and hangers under the constraints of the support and hanger design based on the pipeline BIM data and the main structure BIM data to obtain a first support and hanger design scheme, wherein the first support and hanger design scheme includes multiple first design positions and multiple first design parameters of the supports and hangers. The fitness analysis module is used to perform fitness analysis calculations on the first support bracket design scheme based on the environmental BIM model to obtain the fitness of the first support bracket. The fitness calculation is performed by predicting the first stability and the first material usage of the first support bracket design scheme. The optimization module is used to optimize the support and hanger design scheme to obtain the optimal support and hanger scheme as the design result. The optimization is performed by setting an optimization cluster mechanism based on the constraint strength.

Images