Method for generating engineering real quantity based on BIM (Building Information Modeling)

Abstract

The invention relates to the technical field of building information models, and discloses a BIM (Building Information Modeling)-based engineering physical quantity generation method, which comprises the following steps of S1, importing a BIM model into a system to perform preliminary data processing; s2, performing format standardization processing on the imported data by using a unified standardization process, and for BIM data from different sources, performing automatic matching and similar component and material merging by using a data integration tool; and S3, carrying out standardization processing on the geometric information to sort and standardize material attributes, classifying and combining similar materials, and executing multi-level conflict detection. Through more accurate real-object quantity calculation and automatic data processing, the system optimizes project budgeting and cost management, resource waste and budgeting hyper-branched caused by estimation errors are reduced, meanwhile, the work efficiency is improved through intelligent report generation and real-time data monitoring functions, and project management is more accurate and efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of building information modeling, and in particular to a method for generating engineering physical quantities based on BIM. Background Art

[0002] In traditional engineering project management, the calculation and generation of physical quantities are the core elements that affect key links such as project budget, construction plan, and resource management. However, traditional methods rely on manual calculations or CAD software and lack systematic data integration and intelligent support, resulting in many deficiencies in physical quantity calculation, as shown below:

[0003] Traditional physical quantity calculation methods usually rely on manual estimation or geometric calculations, which are difficult to accurately capture the details of complex engineering structures, resulting in deviations between actual construction resource requirements and budget estimates, which can easily lead to budget overruns and resource waste. In addition, manual calculations are time-consuming, inefficient, and easily affected by human factors, which further exacerbates calculation errors.

[0004] In traditional engineering projects, the lag of physical quantity data makes it difficult to formulate construction plans and effectively allocate resources. It is difficult for managers to obtain accurate data in a timely manner to adjust the construction progress or allocate resources, resulting in construction delays and waste of resources.

[0005] The traditional method of generating physical quantities lacks a mechanism for automatic verification and consistency checking, which can easily lead to data errors and affect construction quality. In addition, the lack of a systematic logging and data review mechanism affects the trust of all parties in the project in the data, which in turn affects the overall cooperation efficiency of the project.

[0006] Therefore, those skilled in the art provide a method for generating engineering physical quantities based on BIM to solve the problems raised in the above background technology. Summary of the invention

[0007] In view of the deficiencies in the prior art, the present invention provides a method for generating engineering physical quantities based on BIM to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for generating engineering physical quantities based on BIM, comprising the following steps:

[0009] S1. Import the BIM model into the system for preliminary data processing;

[0010] S2. Use a unified standardized process to normalize the format of the imported data, and use data integration tools to automatically match and merge similar components and materials for BIM data from different sources;

[0011] S3, standardize the geometric information to organize and standardize material properties, classify and merge similar materials, perform multi-level conflict detection, use intelligent repair algorithms to automatically evaluate the best conflict repair solution, and record repair logs;

[0012] S4. Use adaptive intelligent classification algorithms to classify components and materials according to the properties of components, and provide classification adjustment options based on project characteristics and user needs, allowing users to customize classification;

[0013] S5. According to the architectural design specifications and project requirements, the calculation rules for the physical quantity are set. According to the set quantification rules, the system automatically extracts the relevant component and material information from the BIM model to calculate the physical quantity, and the system automatically records and analyzes the data during the calculation process;

[0014] S6. Automatically verify the generated physical quantity data, identify and correct potential calculation errors, and integrate user feedback functions to collect user feedback and incorporate the feedback into the system through a self-learning mechanism;

[0015] S7. Provides a variety of report templates, users can customize the report format, supports multiple output formats, and adds data visualization tools to the report to automatically review the generated report.

[0016] Preferably, in the step S1, the system is an integrated BIM data processing and engineering physical quantity generation platform, which aims to improve the accuracy and efficiency of physical quantity calculation of engineering projects through automation and intelligent technology. The system can handle complex BIM models, covering the entire process from data import, standardized processing, intelligent classification to physical quantity calculation and report generation.

[0017] Preferably, in the step S2, the algorithm formula for format normalization of the imported data and automatic matching and merging of multi-source data is as follows:

[0018] When normalizing BIM model data, the Z-score standardization method is used to make the data conform to a unified standard format:

[0019]

[0020] Where X is the original data value, μ is the mean of the data, and σ is the standard deviation of the data;

[0021] In multi-source data fusion, cosine similarity is used to judge the similarity of components or materials in different BIM models, and automatic matching and merging are achieved. The cosine similarity formula is:

[0022]

[0023] Among them, A and B are vectors, representing the characteristic vectors of components or materials from different sources, ||A|| and ||B|| represent the norms of vectors A and B;

[0024] Accurately classify and merge material properties. Use the K-means clustering algorithm to automatically classify materials and merge similar materials. The K-means clustering formula is:

[0025] Initial centroid: K random groups of initial centroids μ1, μ2, ..., μ K ;

[0026] Classification: For each data point x i , calculate its distance to each centroid:

[0027]

[0028] Update the centroid: Recalculate the centroid of each cluster, which is the average value of all points in the cluster:

[0029]

[0030] Iteration: Repeat the steps of classification and updating centroids until the centroids no longer change.

[0031] Preferably, in step S2, in order to ensure the consistency of the format and content of data from different sources, a similarity coefficient is used to measure the similarity between data sets to perform a consistency check. The similarity coefficient formula is:

[0032]

[0033] Where A and B are data sets, |A∩B| is the number of elements in the intersection of the data sets, and |A∪B| is the number of elements in the union of the data sets;

[0034] When merging similar components and materials, a weighted average strategy is used to ensure the rationality of the merging results. The weighted average formula is:

[0035]

[0036] Among them, x i are the data to be merged, w i is the corresponding weight and n is the number of data items.

[0037] Preferably, in the step S3, the standardized processing of geometric information and material properties, the application of multi-level conflict detection and intelligent repair algorithm are as follows:

[0038] When repairing geometric conflicts, use the constraint optimization method to find the best repair solution. The constraint optimization formula is:

[0039] Minimizef(x)subject to g i (x)≤0,h j (x) = 0,

[0040] Among them, f(x) is the objective function, g i (x)≤0 is an inequality constraint, h j (x) = 0 is an equality constraint;

[0041] In the standardization of material properties, principal component analysis is used to reduce the data dimension. The principal component analysis dimensionality reduction formula is:

[0042] Z = XW,

[0043] Among them, X is the original data matrix, W is the projection matrix, and Z is the data matrix after dimensionality reduction;

[0044] When generating and recording repair logs, a log probability model is used to quantify the uncertainty in the repair process. The log probability formula is:

[0045]

[0046] Among them, P(log) represents the overall probability of repairing the log, x i is the status or result of each repair operation, and θ is the parameter of the repair model.

[0047] Preferably, in step S4, an adaptive intelligent classification algorithm is used to classify components and materials, and classification adjustment options are provided according to project characteristics and user needs. The following is an example of the algorithm formula:

[0048] To provide more flexible classification adjustment options, the system uses a hierarchical clustering algorithm to cluster components and materials in a hierarchical manner. The hierarchical clustering distance formula is:

[0049]

[0050] Where x and y are the feature vectors of the component or material, n is the dimension of the feature, and d(x,y) represents the Euclidean distance between x and y;

[0051] In order to meet the needs of different projects and users, the classification rules are adjusted using Bayesian optimization. The optimal classification scheme is queried by continuously adjusting the classification parameters. The Bayesian optimization formula is:

[0052]

[0053] Where D is the set of existing data points, is a Gaussian process, with the mean function μ(x) and covariance function k(x,x′) defined at each point x;

[0054] The goal is to find a new point x that maximizes f(x), usually using the expected improvement criterion:

[0055]

[0056] Among them, x + is the location of the optimal solution;

[0057] To further improve the classification effect, the system uses reinforcement learning to continuously adjust the classification strategy based on user feedback and the reinforcement learning value function:

[0058]

[0059] Among them, s is the system state, π is the strategy, r t+1 is the reward at time t+1, and γ is the discount factor.

[0060] Preferably, in the step S5, the system sets the calculation rules of the physical quantity according to the building design specifications and project requirements, and automatically extracts the relevant components and material information from the BIM model to calculate the physical quantity. The algorithm formula is as follows:

[0061] When setting the physical quantity calculation rules, parametric modeling is used to dynamically generate calculation rules. The calculation methods of various components are defined based on the input building specifications and project requirements. The parametric modeling formula is:

[0062] Q=f(p1,p2,...,p n ),

[0063] Among them, Q is the physical quantity to be calculated, p1, p2, ..., p n is a parameter related to the component or material, and f is a parameterized function;

[0064] When automatically extracting relevant component and material information from the BIM model, the feature extraction algorithm can be used to extract key geometry and attribute information for physical quantity calculation. The feature extraction formula is:

[0065] F=φ(G,M),

[0066] Among them, F is the extracted feature set, G is the geometric information of the component, M is the material property information, and φ is the feature extraction function;

[0067] When dealing with complex geometries, three-dimensional integrals are used to calculate their volumes or other related physical quantities. The three-dimensional integral formula is:

[0068] V=∫∫∫ Ω dV,

[0069] Among them, Ω is the three-dimensional space area occupied by the geometric body, and dV is the volume element;

[0070] If the geometry is parametric, the volume integral can be transformed into a double integral over the parameter space:

[0071]

[0072] Where S is the region of parameter space, r(u,v) is the parameterized representation of the geometry, and is the partial derivative with respect to parameters u and v, and × represents the vector cross product.

[0073] Preferably, in the step S6, when automatically checking the generated physical quantity data, a statistical control chart can be used to monitor the calculation results and identify abnormal values ​​or errors. The statistical control chart formula is:

[0074] Suppose that a certain physical quantity X is calculated multiple times, the average of the calculation results is And the standard deviation σ can be expressed as:

[0075]

[0076] Based on the mean and standard deviation, control limits can be defined to monitor the data:

[0077]

[0078] Where UCL and LCL are the upper and lower control limits, and k is the control limit factor.

[0079] Preferably, in step S6, the system automatically verifies the generated physical quantity data, identifies and corrects potential calculation errors, and collects user feedback by integrating the user feedback function, so as to integrate the feedback into the system through the self-learning mechanism. The following is an example of the algorithm formula:

[0080] In order to effectively utilize user feedback, sentiment analysis algorithms can be used to analyze the emotional tendencies of user feedback and decide how to adjust the calculation logic of the system. The sentiment analysis formula is:

[0081] Sentiment analysis can be performed using the bag-of-words model and the TF-IDF weighted method. Assume that the user feedback text is represented by a word set T = {t1, t2, ..., t m}, the TF-IDF value w of each word i It is expressed as:

[0082] TF-IDF(t i ,d,D)=TF(t i ,d)×IDF(t i ,D),

[0083]

[0084] in, It is the word t i frequency in document d;

[0085]

[0086] Where |D| is the total number of documents, |{d∈D:t i ∈d}| is the word containing the word t i The number of documents;

[0087] In order to allow the system to adjust itself according to user feedback, the Q learning algorithm in reinforcement learning can be used to allow the system to continuously improve the calculation rules. The Q learning algorithm formula is:

[0088] The core of Q learning is to update the state-action value function, the formula is:

[0089]

[0090] Among them, s is the current state, a is the action performed in state s, r is the reward obtained after performing action a, s′ is the new state after performing the action, α is the learning rate, and γ is the discount factor.

[0091] Preferably, in step S7, in order to ensure the accuracy of the report content, the system can use a text similarity algorithm in natural language processing to automatically review the text content in the report. The text similarity calculation formula is:

[0092] Assuming there are texts T1 and T2 in the report, the BERT model can encode the text into vectors V1 and V2 and calculate the cosine similarity:

[0093]

[0094] Wherein, V1 and V2 represent the dot product of vectors, and ||V1|| and ||V2|| represent the norms of vectors V1 and V2. The present invention provides a method for generating engineering physical quantities based on BIM. It has the following beneficial effects:

[0095] 1. The present invention optimizes project budgeting and cost management through more accurate physical quantity calculation and automated data processing, thereby reducing resource waste and budget overruns caused by estimation errors. At the same time, intelligent report generation and real-time data monitoring functions improve work efficiency, making project management more accurate and efficient.

[0096] 2. The present invention helps project managers optimize construction progress and resource allocation and reduce the risk of construction delays and resource waste through the precise physical quantity data and adaptive classification functions provided by the system. Its intelligent support improves the effectiveness of construction planning and resource management and makes project operations smoother.

[0097] 3. The present invention ensures the accuracy of physical quantity data and reduces quality problems in construction through automatic verification, error correction and data consistency verification functions. In addition, the system's logging and automatic review functions enhance the traceability and transparency of data and improve the trust and cooperation efficiency of all parties involved in the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0099] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0100] The present invention is described in detail below in conjunction with the accompanying drawings:

[0101] Example:

[0102] Please see attached Figure 1 ,The embodiment of the present invention provides a method for generating engineering physical quantities based on BI M, including the following steps: S1, importing the BI M model into the system to perform preliminary data processing;

[0103] S2. Use a unified standardized process to normalize the format of the imported data, and use data integration tools to automatically match and merge similar components and materials for BIM data from different sources;

[0104] S3, standardize the geometric information to organize and standardize material properties, classify and merge similar materials, perform multi-level conflict detection, use intelligent repair algorithms to automatically evaluate the best conflict repair solution, and record repair logs;

[0105] S4. Use adaptive intelligent classification algorithms to classify components and materials according to the properties of components, and provide classification adjustment options based on project characteristics and user needs, allowing users to customize classification;

[0106] S5. According to the architectural design specifications and project requirements, the calculation rules for the physical quantity are set. According to the set quantification rules, the system automatically extracts the relevant component and material information from the BIM model to calculate the physical quantity, and the system automatically records and analyzes the data during the calculation process;

[0107] S6. Automatically verify the generated physical quantity data, identify and correct potential calculation errors, and integrate user feedback functions to collect user feedback and incorporate the feedback into the system through a self-learning mechanism;

[0108] S7. Provides a variety of report templates, users can customize the report format, supports multiple output formats, and adds data visualization tools to the report to automatically review the generated report.

[0109] Benefits of S1: Improve the accuracy and consistency of data, avoid calculation errors or delays caused by incomplete or erroneous data in the subsequent stages, and ensure the smooth progress of the process; Benefits of S2: Enhance the system's compatibility and processing capabilities for multi-source data, avoid data conflicts or duplications, improve the efficiency and accuracy of data processing, and enable the system to handle complex engineering projects more efficiently; Benefits of S3: Ensure the accuracy and consistency of model data, reduce conflicts and errors in the model, provide a reliable data basis for subsequent physical quantity calculations, and provide clear repair tracking through log records; Benefits of S4: Improve the flexibility and adaptability of the system, and be able to adjust according to specific project conditions to meet the needs of diverse users Benefits of S5: Achieve high efficiency and accuracy in physical quantity calculation, ensure the accuracy of project budget and resource allocation, avoid waste of resources and budget overruns, and provide data support for the smooth progress of the project; Benefits of S6: Reduce the risk of data errors, improve the reliability of calculation results, and enhance the intelligence level of the system, so that it can continue to learn and optimize, and further improve the quality and efficiency of project management; Benefits of S7: Improve the flexibility and customizability of report generation. Users can choose the appropriate report format according to project requirements. Data visualization tools make report content easier to understand, and automatic review functions ensure the accuracy and consistency of reports, thereby enhancing the trust and collaboration efficiency of all parties involved in the project.

[0110] In step S1, the system is an integrated BIM data processing and engineering physical quantity generation platform, which aims to improve the accuracy and efficiency of physical quantity calculation of engineering projects through automation and intelligent technology. The system can handle complex BIM models, covering the entire process from data import, standardized processing, intelligent classification to physical quantity calculation and report generation.

[0111] The system can handle complex BIM model data, covering the entire process from data import to physical quantity calculation and report generation, which means that the system can handle all types of building information models, regardless of their size and complexity, and can effectively handle them to ensure data integrity and consistency;

[0112] The system uses automation technology to seamlessly import BIM model data for preliminary analysis and processing, which reduces the time and potential errors of manual operations, allowing data to quickly enter the standardized processing stage, laying the foundation for subsequent processing steps;

[0113] The system uses intelligent technology to improve the accuracy of physical quantity calculations through standardized processing and intelligent classification. The system's automated processes and intelligent algorithms can quickly process and analyze scale data to ensure the accuracy of calculation results, thereby optimizing project budgeting and cost management.

[0114] In step S2, the imported data is formatted and automatically matched and merged with multi-source data. The following is the algorithm formula:

[0115] When normalizing BIM model data, the Z-score standardization method is used to make the data conform to a unified standard format:

[0116]

[0117] Where X is the original data value, μ is the mean of the data, and σ is the standard deviation of the data;

[0118] In multi-source data fusion, cosine similarity is used to judge the similarity of components or materials in different BIM models, and automatic matching and merging are achieved. The cosine similarity formula is:

[0119]

[0120] Among them, A and B are vectors, representing the characteristic vectors of components or materials from different sources, ||A|| and ||B|| represent the norms of vectors A and B;

[0121] Accurately classify and merge material properties. Use the K-means clustering algorithm to automatically classify materials and merge similar materials. The K-means clustering formula is:

[0122] Initial centroid: K random groups of initial centroids μ1, μ2, ..., μ K ;

[0123] Classification: For each data point x i , calculate its distance to each centroid:

[0124]

[0125] Update the centroid: Recalculate the centroid of each cluster, which is the average value of all points in the cluster:

[0126]

[0127] Iteration: Repeat the steps of classification and updating centroids until the centroids no longer change.

[0128] Z-score standardization processing: By standardizing the data, BIM model data from different sources conform to a unified standard format. This operation converts the original data into data with the same dimension, eliminates the scale differences between data from different sources, makes the comparison and analysis between different data sets more intuitive, reduces the calculation errors caused by data distribution differences, and improves the accuracy and reliability of data processing;

[0129] Cosine similarity is used for automatic matching and merging of multi-source data. In the integration of multi-source data, cosine similarity is used to identify and match similar components or materials, and automatically merge similar items. The process reduces data redundancy and ensures the accuracy and consistency of multi-source data integration.

[0130] The K-means clustering algorithm is used to classify and merge material properties by calculating the distance between each data point and the centroid and continuously updating the position of the centroid. The K-means clustering algorithm optimizes the classification results in the iterative process until the centroid remains stable. Its optimization mechanism ensures the accuracy of the classification results, enables the same type of materials to be correctly classified, reduces data redundancy, improves data processing efficiency, and ensures the consistency of material information in physical quantity calculations.

[0131] In step S2, in order to ensure the consistency of format and content of data from different sources, a similarity coefficient is used to measure the similarity between data sets and perform a consistency check. The formula for the similarity coefficient is:

[0132]

[0133] Where A and B are data sets, |A∩B| is the number of elements in the intersection of the data sets, and |A∪B| is the number of elements in the union of the data sets;

[0134] When merging similar components and materials, a weighted average strategy is used to ensure the rationality of the merging results. The weighted average formula is:

[0135]

[0136] Among them, x i are the data to be merged, w i is the corresponding weight and n is the number of data items.

[0137] The similarity coefficient is used to measure the similarity between data sets. The similarity measurement method can effectively detect the consistency of the content of BIM data from different sources. In the data processing process, the similarity coefficient helps to identify and verify whether the BIM data from different sources are consistent with each other, ensuring that the differences between data sets are minimized and improving the accuracy and consistency of data integration;

[0138] Weighted average is used to merge similar components and materials to ensure that when merging similar components or materials, more important or more reliable data has a greater impact on the merged results. The strategy avoids the errors caused by simple averaging and ensures the rationality and accuracy of the merged results. By setting weights reasonably, the system can balance the differences between different data sources during the merging process, obtain merged results that conform to the actual situation, effectively avoid data bias, and improve the reliability of the merged results.

[0139] In step S3, the standardization of geometric information and material properties, multi-level conflict detection and intelligent repair algorithm are applied. The algorithm formula is as follows:

[0140] When repairing geometric conflicts, use the constraint optimization method to find the best repair solution. The constraint optimization formula is:

[0141] Minimizef(x)subject to g i (x)≤0,h j (x) = 0,

[0142] Among them, f(x) is the objective function, g i (x)≤0 is an inequality constraint, h j (x) = 0 is an equality constraint;

[0143] In the standardization of material properties, principal component analysis is used to reduce the data dimension. The principal component analysis dimensionality reduction formula is:

[0144] Z = XW,

[0145] Among them, X is the original data matrix, W is the projection matrix, and Z is the data matrix after dimensionality reduction;

[0146] When generating and recording repair logs, a log probability model is used to quantify the uncertainty in the repair process. The log probability formula is:

[0147]

[0148] Among them, P(log) represents the overall probability of repairing the log, x i is the status or result of each repair operation, and θ is the parameter of the repair model.

[0149] Constrained Optimization By defining the objective function and constraints, the system can search for the best solution among multiple possible repair schemes to ensure that the geometric conflict is repaired while still meeting the design requirements and physical constraints of the project. The design of the objective function usually considers the balance between the cost and effect of the repair, and achieves the most economical and effective conflict repair.

[0150] Principal component analysis reduces the data dimension, extracts the main features of material properties, removes redundant and unimportant information, makes the data more concise, easier to process and analyze, and retains the main information of the data;

[0151] The log probability model quantifies the uncertainty in the repair operation by calculating the overall probability of the repair process. Quantification can help the system better understand and record the risks and reliability of each repair step, further improving the reliability and effectiveness of the repair process.

[0152] In step S4, an adaptive intelligent classification algorithm is used to classify components and materials, and classification adjustment options are provided according to project characteristics and user needs. The following is an example of the algorithm formula:

[0153] To provide more flexible classification adjustment options, the system uses a hierarchical clustering algorithm to cluster components and materials in a hierarchical manner. The hierarchical clustering distance formula is:

[0154]

[0155] Where x and y are the feature vectors of the component or material, n is the dimension of the feature, and d(x,y) represents the Euclidean distance between x and y;

[0156] In order to meet the needs of different projects and users, the classification rules are adjusted using Bayesian optimization. The optimal classification scheme is queried by continuously adjusting the classification parameters. The Bayesian optimization formula is:

[0157]

[0158] Where D is the set of existing data points, is a Gaussian process, with the mean function μ(x) and covariance function k(x,x′) defined at each point x;

[0159] The goal is to find a new point x that maximizes f(x), usually using the expected improvement criterion:

[0160]

[0161] Among them, x + is the location of the optimal solution;

[0162] To further improve the classification effect, the system uses reinforcement learning to continuously adjust the classification strategy based on user feedback and the reinforcement learning value function:

[0163]

[0164] Among them, s is the system state, π is the strategy, r t+1 is the reward at time t+1, and γ is the discount factor.

[0165] Through the combined application of hierarchical clustering, Bayesian optimization and reinforcement learning, the system in step S4 can achieve flexible classification adjustment and precise optimization of classification rules. Hierarchical clustering provides users with multi-level classification options, Bayesian optimization ensures the optimal classification rules, and reinforcement learning enables the system to adaptively adjust the classification strategy based on user feedback. The synergistic effect of the algorithms significantly improves the adaptability, flexibility and intelligence of the system in engineering projects, ensuring the classification needs in project management.

[0166] In step S5, the system sets the calculation rules for physical quantities according to the architectural design specifications and project requirements, and automatically extracts relevant component and material information from the BIM model for physical quantity calculation. The algorithm formula is as follows:

[0167] When setting the physical quantity calculation rules, parametric modeling is used to dynamically generate calculation rules. The calculation methods of various components are defined based on the input building specifications and project requirements. The parametric modeling formula is:

[0168] Q=f(p1,p2,...,p n ),

[0169] Among them, Q is the physical quantity to be calculated, p1, p2, ..., p n is a parameter related to the component or material, and f is a parameterized function;

[0170] When automatically extracting relevant component and material information from the BIM model, the feature extraction algorithm can be used to extract key geometry and attribute information for physical quantity calculation. The feature extraction formula is:

[0171] F=φ(G,M),

[0172] Among them, F is the extracted feature set, G is the geometric information of the component, M is the material property information, and φ is the feature extraction function;

[0173] When dealing with complex geometries, three-dimensional integrals are used to calculate their volumes or other related physical quantities. The three-dimensional integral formula is:

[0174] V=∫∫∫ Ω dV,

[0175] Among them, Ω is the three-dimensional space area occupied by the geometric body, and dV is the volume element;

[0176] If the geometry is parametric, the volume integral can be transformed into a double integral over the parameter space:

[0177]

[0178] Where S is the region of parameter space, r(u,v) is the parameterized representation of the geometry, and is the partial derivative with respect to parameters u and v, and × represents the vector cross product.

[0179] Through the combination of parametric modeling, feature extraction and three-dimensional integration algorithms, the system in step S5 can flexibly set physical quantity calculation rules according to architectural design specifications and project requirements, automatically extract relevant information from the BIM model, and accurately process the physical quantity of complex geometric bodies. The application of the algorithm ensures the efficiency, accuracy and adaptability of physical quantity calculation, providing a solid data foundation for the budget preparation, construction plan and resource management of engineering projects.

[0180] In step S6, when automatically checking the generated physical quantity data, a statistical control chart can be used to monitor the calculation results and identify abnormal values ​​or errors. The statistical control chart formula is:

[0181] Suppose that a certain physical quantity X is calculated multiple times, the average of the calculation results is And the standard deviation σ can be expressed as:

[0182]

[0183] Based on the mean and standard deviation, control limits can be defined to monitor the data:

[0184]

[0185] Where UCL and LCL are the upper and lower control limits, and k is the control limit factor.

[0186] By using statistical control charts and their related formulas, the system in step S6 can effectively monitor the stability of the physical quantity calculation data, identify outliers and errors in a timely manner, and take corresponding measures to correct them. The monitoring mechanism ensures the reliability and accuracy of the data, provides strong guarantees for the physical quantity calculation of engineering projects, and supports the continuous improvement of projects.

[0187] In step S6, the system automatically verifies the generated physical quantity data, identifies and corrects potential calculation errors, and collects user feedback by integrating the user feedback function, so as to integrate the feedback into the system through the self-learning mechanism. The following is an example of the algorithm formula:

[0188] In order to effectively utilize user feedback, sentiment analysis algorithms can be used to analyze the emotional tendencies of user feedback and decide how to adjust the calculation logic of the system. The sentiment analysis formula is:

[0189] Sentiment analysis can be performed using the bag-of-words model and the TF-IDF weighted method. Assume that the user feedback text is represented by a word set T = {t1, t2, ..., t m}, the TF-IDF value w of each word i It is expressed as:

[0190] TF-IDF(t i ,d,D)=TF(t i ,d)×IDF(t i ,D),

[0191]

[0192] in, It is the word t i frequency in document d;

[0193]

[0194] Where |D| is the total number of documents, |{d∈D:t i ∈d}| is the word containing the word t i The number of documents;

[0195] In order to allow the system to adjust itself according to user feedback, the Q learning algorithm in reinforcement learning can be used to allow the system to continuously improve the calculation rules. The Q learning algorithm formula is:

[0196] The core of Q learning is to update the state-action value function, the formula is:

[0197]

[0198] Among them, s is the current state, a is the action performed in state s, r is the reward obtained after performing action a, s′ is the new state after performing the action, α is the learning rate, and γ is the discount factor.

[0199] Through sentiment analysis and Q-learning algorithms, the system in step S6 can effectively analyze user feedback, identify potential problems, and continuously optimize the calculation rules through self-adjustment. Sentiment analysis enables the system to understand user needs and optimize accordingly, while Q-learning provides the system with the ability of continuous learning and adaptive adjustment. The combination of algorithms improves the intelligence level of the system and ensures that the accuracy of physical quantity calculation and user satisfaction continue to improve.

[0200] In step S7, to ensure the accuracy of the report content, the system can use the text similarity algorithm in natural language processing to automatically review the text content in the report. The text similarity calculation formula is:

[0201] Assuming there are texts T1 and T2 in the report, the BERT model can encode the text into vectors V1 and V2 and calculate the cosine similarity:

[0202]

[0203] Among them, V1 and V2 represent the dot product of vectors, and ||V1|| and ||V2|| represent the norms of vectors V1 and V2.

[0204] Through the cosine similarity calculation formula based on the BERT model, the system in step S7 can effectively review the report content and ensure the consistency and accuracy of the text. The application of this algorithm improves the degree of automation in the report generation process, reduces human errors, and ultimately improves the quality and reliability of the report, providing more reliable document support for project management.

[0205] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for generating engineering physical quantities based on BIM, characterized in that: The steps include: S1. Import the BIM model into the system for preliminary data processing; S2. Use a unified standardized process to normalize the format of the imported data, and use data integration tools to automatically match and merge similar components and materials for BIM data from different sources; S3, standardize the geometric information to organize and standardize material properties, classify and merge similar materials, perform multi-level conflict detection, use intelligent repair algorithms to automatically evaluate the best conflict repair solution, and record repair logs; S4. Use adaptive intelligent classification algorithms to classify components and materials according to the properties of components, and provide classification adjustment options based on project characteristics and user needs, allowing users to customize classification; S5. According to the architectural design specifications and project requirements, the calculation rules for the physical quantity are set. According to the set quantification rules, the system automatically extracts the relevant component and material information from the BIM model to calculate the physical quantity, and the system automatically records and analyzes the data during the calculation process; S6. Automatically verify the generated physical quantity data, identify and correct potential calculation errors, and integrate user feedback functions to collect user feedback and incorporate the feedback into the system through a self-learning mechanism; S7. Provides a variety of report templates, users can customize the report format, supports multiple output formats, and adds data visualization tools to the report to automatically review the generated report.

2. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In the S1 step, the system is an integrated BIM data processing and engineering physical quantity generation platform, which aims to improve the accuracy and efficiency of physical quantity calculation of engineering projects through automation and intelligent technology. The system can handle complex BIM models, covering the entire process from data import, standardized processing, intelligent classification to physical quantity calculation and report generation.

3. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In the step S2, the algorithm formula is used to normalize the format of the imported data and automatically match and merge multi-source data. When normalizing BIM model data, the Z-score standardization method is used to make the data conform to a unified standard format: Where X is the original data value, μ is the mean of the data, and σ is the standard deviation of the data; In multi-source data fusion, cosine similarity is used to judge the similarity of components or materials in different BIM models, and automatic matching and merging are achieved. The cosine similarity formula is: Among them, A and B are vectors, representing the characteristic vectors of components or materials from different sources, and ∥A∥ and ∥B∥ represent the norms of vectors A and B; Accurately classify and merge material properties. Use the K-means clustering algorithm to automatically classify materials and merge similar materials. The K-means clustering formula is: Initial centroid: K random groups of initial centroids μ1, μ2, ..., μ K ; Classification: For each data point x i , calculate its distance to each centroid: Update the centroid: Recalculate the centroid of each cluster, which is the average value of all points in the cluster: Iteration: Repeat the steps of classification and updating centroids until the centroids no longer change.

4. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In step S2, in order to ensure the consistency of format and content of data from different sources, a similarity coefficient is used to measure the similarity between data sets and perform a consistency check. The similarity coefficient formula is: Where A and B are data sets, |A∩B| is the number of elements in the intersection of the data sets, and |A∪B| is the number of elements in the union of the data sets; When merging similar components and materials, a weighted average strategy is used to ensure the rationality of the merging results. The weighted average formula is: Among them, x i are the data to be merged, w i is the corresponding weight and n is the number of data items.

5. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In the S3 step, the standardization of geometric information and material properties, the application of multi-level conflict detection and intelligent repair algorithms are as follows: When repairing geometric conflicts, use the constraint optimization method to find the best repair solution. The constraint optimization formula is: Minimizef(x)subject to g i (x)≤0,h j (x)=0, Among them, f(x) is the objective function, g i (x)≤0 is an inequality constraint, h j (x) = 0 is an equality constraint; In the standardization of material properties, principal component analysis is used to reduce the data dimension. The principal component analysis dimensionality reduction formula is: Z = XW, Among them, X is the original data matrix, W is the projection matrix, and Z is the data matrix after dimensionality reduction; When generating and recording repair logs, a log probability model is used to quantify the uncertainty in the repair process. The log probability formula is: Among them, P(log) represents the overall probability of repairing the log, x i is the status or result of each repair operation, and θ is the parameter of the repair model.

6. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In the step S4, an adaptive intelligent classification algorithm is used to classify components and materials, and classification adjustment options are provided according to project characteristics and user needs. The following is an example of the algorithm formula: To provide more flexible classification adjustment options, the system uses a hierarchical clustering algorithm to cluster components and materials in a hierarchical manner. The hierarchical clustering distance formula is: Where x and y are the feature vectors of the component or material, n is the dimension of the feature, and d(x,y) represents the Euclidean distance between x and y; In order to meet the needs of different projects and users, the classification rules are adjusted using Bayesian optimization. The optimal classification scheme is queried by continuously adjusting the classification parameters. The Bayesian optimization formula is: Where D is the set of existing data points, is a Gaussian process, with the mean function μ(x) and covariance function k(x,x′) defined at each point x; The goal is to find a new point x that maximizes f(x), usually using the expected improvement criterion: Among them, x + is the location of the optimal solution; To further improve the classification effect, the system uses reinforcement learning to continuously adjust the classification strategy based on user feedback and the reinforcement learning value function: Among them, s is the system state, π is the strategy, r t+1 is the reward at time t+1, and γ is the discount factor.

7. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In the step S5, the system sets the calculation rules for the physical quantity according to the architectural design specifications and project requirements, and automatically extracts the relevant components and material information from the BIM model to calculate the physical quantity. The algorithm formula is as follows: When setting the physical quantity calculation rules, parametric modeling is used to dynamically generate calculation rules. The calculation methods of various components are defined based on the input building specifications and project requirements. The parametric modeling formula is: Q=f(p1,p2,...,p n ), Among them, Q is the physical quantity to be calculated, p1, p2, ..., p n is a parameter related to the component or material, and f is a parameterized function; When automatically extracting relevant component and material information from the BIM model, the feature extraction algorithm can be used to extract key geometry and attribute information for physical quantity calculation. The feature extraction formula is: F=φ(G,M), Among them, F is the extracted feature set, G is the geometric information of the component, M is the material property information, and φ is the feature extraction function; When dealing with complex geometries, three-dimensional integrals are used to calculate their volumes or other related physical quantities. The three-dimensional integral formula is: V=∫∫∫ Ω dV, Among them, Ω is the three-dimensional space area occupied by the geometric body, and dV is the volume element; If the geometry is parametric, the volume integral can be transformed into a double integral over the parameter space: Where S is the region of parameter space, r(u,v) is the parameterized representation of the geometry, and is the partial derivative with respect to parameters u and v, and × represents the vector cross product.

8. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In step S6, when automatically checking the generated physical quantity data, a statistical control chart can be used to monitor the calculation results and identify abnormal values ​​or errors. The statistical control chart formula is: Suppose that a certain physical quantity X is calculated multiple times, the average of the calculation results is And the standard deviation σ can be expressed as: Based on the mean and standard deviation, control limits can be defined to monitor the data: Where UCL and LCL are the upper and lower control limits, and k is the control limit factor.

9. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In step S6, the system automatically verifies the generated physical quantity data, identifies and corrects potential calculation errors, and collects user feedback by integrating the user feedback function, so as to integrate the feedback into the system through the self-learning mechanism. The following is an example of the algorithm formula: In order to effectively utilize user feedback, sentiment analysis algorithms can be used to analyze the emotional tendencies of user feedback and decide how to adjust the system's calculation logic. The sentiment analysis formula is: Sentiment analysis can be performed using the bag-of-words model and the TF-IDF weighted method. Assume that the user feedback text is represented by a word set T = {t1, t2, ..., t m }, the TF-IDF value w of each word i It is expressed as: TF-IDF(t i ,d,D)=TF(t i ,d)×IDF(t i ,D), in, It is the word t i frequency in document d; Where |D| is the total number of documents, |{d∈D:t i ∈d}| is the word containing the word t i The number of documents; In order to allow the system to adjust itself according to user feedback, the Q learning algorithm in reinforcement learning can be used to allow the system to continuously improve the calculation rules. The Q learning algorithm formula is: The core of Q learning is to update the state-action value function, the formula is: Among them, s is the current state, a is the action performed in state s, r is the reward obtained after performing action a, s′ is the new state after performing the action, α is the learning rate, and γ is the discount factor.

10. The method for generating engineering physical quantities based on BIM according to claim 1, characterized in that: In step S7, in order to ensure the accuracy of the report content, the system can use the text similarity algorithm in natural language processing to automatically review the text content in the report. The text similarity calculation formula is: Assuming there are texts T1 and T2 in the report, the BERT model can encode the text into vectors V1 and V2 and calculate the cosine similarity: Among them, V1 and V2 represent the dot product of vectors, and ∥V1∥ and ∥V2∥ represent the norms of vectors V1 and V2.

Images