Fabricated building intelligent drawing method and system based on BIM technology

Abstract

The invention discloses a fabricated building intelligent drawing method and system based on the BIM technology, and relates to the technical field of fabricated buildings and BIM. The method comprises the steps that structural extraction and mapping are conducted on component information, multi-dimensional information of full-professional components of a BIM platform is collected, and information completeness and compliance are verified based on four levels; converting into a machine readable format based on industry specifications, enterprise standards and project requirements, and constructing a parameterized drawing rule base to generate drawing rules; based on the verified component information and drawing output rules, a drawing is automatically generated, and bidirectional association among the drawing, the component and the BIM platform is established; the drawing quality is intelligently verified and dynamically optimized, errors are processed stage by stage, rules are adjusted, the problems that the drawing output efficiency is low and quality control is difficult are effectively solved, and the drawing output efficiency and the drawing output accuracy are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of prefabricated buildings and BIM technology, in particular to a prefabricated building intelligent drawing method and system based on BIM technology. BACKGROUND

[0002] With the deepening of the building industrialization and green building policy, prefabricated buildings have become the core direction of the transformation of the construction industry, and the prefabricated building drawing link based on BIM technology as the key technical carrier connecting the design end and the production and construction end, its drawing efficiency and accuracy directly affect the reliability of the whole process collaboration of design, production and construction and the overall project promotion efficiency; With the gradual application of BIM technology, this field has realized the technical leap from traditional manual CAD drawing to BIM model assisted drawing, which automatically generates preliminary drawings by extracting component basic parameters in BIM model, so that the prefabricated building drawing link preliminarily meets the basic drawing demand of conventional components such as prefabricated wallboards and laminated boards, but the existing technology still has many deficiencies: The traditional technology lacks sufficient quantitative mapping association between BIM model component full parameters and drawing output specifications, the BIM drawing control strategy does not integrate multi-specialty collaborative constraint parameters of architecture, structure and electromechanical engineering, lacks real-time monitoring and dynamic adaptation mechanism of component production process requirements and on-site installation positioning data, and the drawing quality verification does not consider the relevance distribution of drawing information under different component types, resulting in difficulties in continuous generation of drawings for wide component types from standard components to special-shaped components, large drawing errors, and low drawing efficiency and quality stability.

[0003] In order to solve the above-mentioned defects, the present application provides a technical scheme. SUMMARY

[0004] The purpose of the present application is to solve the problems of low drawing efficiency and quality control difficulty, and to provide a prefabricated building intelligent drawing method and system based on BIM technology.

[0005] The purpose of the present application can be realized by the following technical scheme: A prefabricated building intelligent drawing method based on BIM technology, comprising: S1, structured extraction and mapping of component information: in the BIM platform environment, the multi-source information of the prefabricated building platform is structured extracted, and the mapping relationship with the drawing elements is established; S2, parameterized construction of drawing rules: based on prefabricated building design specifications and enterprise standards, a parameterized drawing rule library is constructed in the BIM supporting plug-in; S3, automatic generation and association of drawings: based on the verified complete component information and parameterized drawing rules, standard drawings are automatically generated through the BIM platform, and a bidirectional association relationship between the drawings, components and BIM platform is established. S4. Intelligent verification and dynamic optimization of drawing quality: Based on intelligent algorithms and rule engines, multi-dimensional quality verification of generated drawings is performed, and dynamic optimization and feedback learning are carried out on the problems found.

[0006] As a further improvement of the present invention, the specific implementation process of step S1 is as follows: Obtain the full-discipline component family library of the prefabricated building BIM platform, including precast concrete, steel structure, electromechanical integration and decoration integrated components, and each component is marked with a unique ID; the full-discipline component family library also includes parameters for production, installation and operation and maintenance in all dimensions; By establishing real-time communication with the digital twin of the component through the open API interface of the BIM platform, information on the geometric feature layer, material performance layer, and connection characteristic layer is extracted layer by layer. Dynamic attributes of the component are extracted based on Python scripts and stored in JSON format. A structured dataset is constructed with the component ID as the unique key value. Based on the four dimensions of function, production, installation and operation and maintenance, component information is sorted out and a dynamic adaptation mapping matrix between component parameters and drawing elements is constructed. A four-level verification chain is constructed: Level 1 uses a hash algorithm for real-time verification; Level 2 uses a knowledge graph for correlation verification; Level 3 uses a neural network combined with an industry standard database for integrity verification; and Level 4 calls the industry standard database for compliance verification. After verification, a supplementary list and a component information completeness report are generated.

[0007] As a further improvement of the present invention, the specific operation process of step S2 includes: Construct a three-tiered rule framework encompassing specifications, standards, and projects: National-level rules extract mandatory requirements from industry standard databases, including detailed proportional drawings of horizontal joint nodes in precast wall panels and marking of weld leg dimensions and lengths for steel structure welds; enterprise-level rules refine production adaptability clauses based on enterprise standards; project-level rules supplement clauses based on project-specific needs. Code all rules, label the applicable component types and drawing types, and form a rule list; The textual specifications of the rule list are converted into structured rule expressions using Python scripts; a rule trigger matrix is ​​constructed based on two dimensions: component type and drawing type, and the matrix cells store the corresponding parameterized expressions of the rules. National-level rules are marked in red to set the highest mandatory enforcement authority, enterprise-level rules are marked in yellow to allow adjustment of parameter thresholds, and project-level rules are marked in blue to enable temporary disabling authority; a rule execution engine is developed, which has a built-in rule trigger matrix and a two-way communication channel for component parameters. When a user initiates a drawing command, the rule set is automatically matched. During execution, it is controlled based on hierarchical permissions. Drawing requests that do not pass the mandatory rules are blocked and a rectification list is displayed. At the same time, the rule execution log is recorded to the central database.

[0008] As a further improvement of the present invention, the specific operation process of step S2 further includes: The performance evaluation of rule usage is based on the rule execution log of the BIM platform, using a formula. Obtain the rule applicability index ,in, Indicates the number of times the rule is triggered. Total number of plots The ratio, i.e., the rule trigger frequency, Indicates the first Secondary users adjust the fit; This indicates the total number of times the user manually adjusted the rules after they were triggered. Indicates the first The number of times a rule conflicts with other rules. This represents the total number of conflict events that occur between all rules. Do not assign weighting factors to rule triggering frequency, user adjustment consistency, and rule conflict rate; When the rule applicability index is lower than the preset threshold, the current rule is marked as needing optimization; if it is lower than the preset threshold multiple times and optimization is ineffective, the rule is directly disabled or eliminated. Generate a rule base health report, including rule applicability ranking, conflict rate ranking, a list of rules to be optimized, and optimization suggestions.

[0009] As a further improvement of the present invention, the specific implementation process of step S3 is as follows: Receive the drawing instructions processed by the S2 step rule engine, identify the target component set and drawing type requirements, and determine the drawing scale, viewport range and section position. Layout drawings use a spatial grid algorithm to arrange plan and elevation views and control display priority; detail drawings extract node construction parameters to generate sectional views and enlarged partial views; production drawings call production process data to generate processing drawings. Through formula Obtain drawing priority index ,in, This indicates the inherent weight of the rule; This indicates the urgency of the situation. These are the component production deadline and the current time, respectively. To ensure the completeness of component information, These are the influence weight factors of the inherent weight of the rule, the time urgency, and the completeness of the component information, respectively; The dimensioning layer generates dimension chains and tolerance annotations based on geometric feature data, and oversized components trigger transport direction annotations; the technical annotation layer calls material and connection data to annotate performance parameters and seismic requirements; the symbol annotation layer embeds predefined symbols and adjusts their positions; BIM data is synchronized through hash verification, and annotations are automatically updated when parameters change; Establish a 3D database linking drawings, components, and the platform, and achieve two-way linkage based on lightweight technology; the cloud platform synchronously generates multi-disciplinary drawings and detects conflicts, generates drawing version IDs to bind component versions, and records version logs.

[0010] As a further improvement of the present invention, the specific implementation process of step S4 includes: Based on the parameterized drawing rule library built on S2, the mandatory rule compliance of the generated drawings is scanned; image recognition is used to detect standardization issues, view features are extracted based on convolutional neural networks to verify dimensional deviations, technical description text is parsed based on natural language processing algorithms, and adversarial network training algorithms are introduced to detect line type confusion and legend overlap; the verification algorithm interacts with the drawing data in real time through the BIM platform API interface, and a unique verification ID is generated for each drawing and bound to the component ID; A three-tiered early warning mechanism is established: Level 1 warning blocks non-compliant drawings that could affect structural safety and triggers audible and visual alarms; Level 2 warning generates enterprise-level rule modification suggestions; and Level 3 warning provides project-level rule optimization suggestions. A feedback dataset is constructed based on the verification logs, using the formula... The reward coefficient for a single map output task ,in, This indicates the number of critical errors that occurred during this drawing output task; This indicates the number of general errors generated during this drawing production task.

[0011] As a further improvement of the present invention, the specific implementation process of step S4 further includes: Through formula The cumulative discount reward coefficient is obtained, where, Indicates the first One drawing output task; Indicates the first The discount factor for the reward coefficient of each map-producing task; The optimal strategy is derived based on the cumulative discount reward coefficient; a rule optimization report is generated quarterly, using a formula. The comprehensive index of influence is obtained, among which, This represents the set of component types constrained by the current rule. This represents the weight of component type b in the project. This represents the average drawing frequency of component type b; Based on the verification results, various error distributions are statistically analyzed to generate a professional quality radar chart. The cloud platform shares data and triggers cross-professional verification. The drawings that are subject to mandatory verification generate quality certification labels, which are stored on the blockchain along with the version ID.

[0012] A second aspect of the present invention provides an intelligent drawing system for prefabricated buildings based on BIM technology, comprising: Component Information Processing Module: Obtains the full-discipline component family library from the BIM platform, generates a structured dataset, extracts geometric, material, and connection information layer by layer through API interfaces, obtains dynamic attributes based on Python scripts and stores them in JSON format; constructs a dynamic mapping matrix between components and drawing elements, equipped with a four-level verification chain, and generates a supplementary list and a completeness report; Rule Management Module: Constructs a three-tier rule framework of specifications, standards, and projects; converts rules into parameterized expressions based on Python scripts; builds a trigger matrix; sets permission levels; develops a rule execution engine; and generates health reports. Drawing generation and association module: Parses drawing instructions to generate views, arranges the drawing order based on the drawing priority index; parametrically drives dimensions, technical and symbol annotations; constructs a 3D association database, outputs multi-format drawings and performs blockchain signature; Quality verification and optimization module: Verifies the compliance of drawings based on multiple algorithms, builds a hierarchical early warning mechanism; optimizes rules based on log feedback, adjusts thresholds based on cumulative discount reward coefficients; generates quality reports and certification labels.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs a dynamic adaptation mapping relationship between the full-dimensional information of components and the elements to be drawn. Combined with a four-level verification chain analysis, it quantifies the impact of missing and incorrect component information on the quality of the drawings and generates a component information completeness index and a rule applicability index. Based on the component information completeness index and the rule applicability index, it generates a priority ranking range for drawings. Combined with a multi-professional collaboration mechanism and blockchain anti-tampering technology, it optimizes the drawing generation rate and cross-professional adjustment step size. This effectively solves the problems of low drawing efficiency and difficult quality control caused by information gaps and poor rule adaptability in traditional technologies, and significantly improves the accuracy and efficiency of drawings for prefabricated buildings. Attached Figure Description

[0014] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0016] It should be understood that the terms “comprising” and “including” used in this disclosure and claims indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0017] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this disclosure. As used in this disclosure and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this disclosure and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations.

[0018] like Figure 1 As shown, a BIM-based intelligent drawing method for prefabricated buildings includes structured extraction and mapping of component information, parametric construction of drawing rules, automated generation and association of drawings, and intelligent verification and dynamic optimization of drawing quality.

[0019] S1. Structured Extraction and Mapping of Component Information: In the BIM platform environment, multi-source information of the prefabricated building platform is extracted in a structured manner, and a mapping relationship is established with the drawing elements. The specific implementation process is as follows: The system acquires a comprehensive component family library from the prefabricated building BIM platform, including precast concrete components, steel structure components, electromechanical integrated components, and integrated decoration components. Precast concrete components include wall panels, floor slabs, beams, columns, and stairs; steel structure components include steel columns, steel beams, and supports; and electromechanical integrated components include precast manholes and embedded pipelines. Furthermore, the comprehensive component family library includes full-dimensional information on component production parameters, installation parameters, and operation and maintenance parameters, with each component labeled with a unique ID. By establishing a real-time communication channel with the digital twin of the components through the open API interface of the BIM platform, information can be extracted layer by layer: Geometric feature layer: Extracts the three-dimensional dimensional accuracy of components, surface flatness, three-dimensional coordinates and diameter tolerance of reserved holes, spatial positioning parameters and anchoring depth of embedded parts, and allowable deviation of component assembly gap; Material performance layer: Extracting concrete strength grade, steel reinforcement grade, thermal conductivity of insulation material, and weather resistance grade of waterproof material; Connection characteristics layer: Extracts node connection type, connector specifications, allowable deviation of node gap, and performance parameters of sealing material; Real-time monitoring of BIM platform changes based on Python scripts, extracting dynamic attributes of components, including allowable values ​​for component hoisting angle deviation and minimum safe distance from adjacent components; Data is stored in JSON format, and a structured dataset is built with the component ID as the unique key. At the same time, the dynamic attributes of the digital twin are synchronized in real time through Python scripts to ensure that the information is consistent with the actual state of the component. Component information is categorized and organized based on component function, production parameters, installation parameters, and operation and maintenance parameters: The functional dimension is divided into structural load-bearing, enclosure and partition, and equipment integration; the production parameter dimension marks the casting method of precast concrete components and the processing accuracy level of steel structure components; the installation parameter dimension is associated with the hoisting sequence number and positioning coordinates of components; the operation and maintenance dimension records the component maintenance cycle and the replacement threshold of vulnerable parts. A dynamic adaptation mapping matrix is ​​constructed based on component parameters and drawing elements. The horizontal dimension consists of drawing elements, including view representations, dimension annotations, and technical specifications. The vertical dimension is divided into mapping subsets based on component type. Each subset contains the mapping relationship between core parameters and drawing elements. The mapping relationship within the subset can be automatically adjusted as the project progresses. Based on a pre-trained component information feature library, a four-level verification chain is constructed, including: Level 1 Real-time Verification: The consistency between component parameters and version parameters is compared using a hash algorithm. When the version is modified but the parameters are not updated synchronously, a red alert is triggered and the difference parameters are displayed. Second-level correlation verification: Based on knowledge graph technology, a component parameter correlation network is constructed, and combined with a convolutional neural network to identify erroneous parameters, generate yellow prompts, and recommend correct correlation parameters; Level 3 integrity verification: Predicts missing parameters using an LSTM neural network and provides supplementary suggestions based on an industry standard database; the industry standard database includes the "Technical Standard for Prefabricated Concrete Buildings", the "Construction Quality Acceptance Standard for Steel Structures", and the "Building Modular Coordination Standard". Level 4 Compliance Verification: This involves calling the industry standards database to verify whether the component parameters comply with the standards required by the project location. If non-compliance is found, an orange alert is triggered and the relevant standard clauses are cited. After verification, a priority-based completion list is generated, with key parameters affecting production marked as the highest priority and triggering immediate reminders. Key parameters include geometric accuracy, material properties, connection nodes, and production processes. A component information completeness report is generated, which statistically analyzes the missing rate of core parameters based on architectural, structural, and mechanical and electrical disciplines. When the missing rate of core parameters is >5%, the subsequent drawing process is blocked until the missing parameters are completed. When the missing rate of non-core parameters is ≤10%, drawing can be started first, but it is marked as pending completion, and a reminder for completion is set.

[0020] S2. Parametric Construction of Drawing Rules: Based on prefabricated building design specifications and enterprise standards, a parametric drawing rule library is constructed in the BIM supporting plugin; the specific construction process is as follows: S201. Standards and Specifications Review: Construct a three-tiered rule framework of standards, specifications, and projects, clarifying the weight and scope of application of drawing requirements at each level. National-level rules: Mandatory drawing requirements are extracted from the industry standard library. Mandatory drawing requirements include that the horizontal joint nodes of precast wall panels must be detailed at a 1:2 scale and that the weld annotations of steel structure components must include the weld leg size and length. Enterprise-level rule layer: Combine enterprise standards to refine production adaptability clauses, such as requiring the weld shear strength grade to be marked on the reinforcement layout drawing of composite floor slab trusses, or requiring prefabricated bathroom units to be accompanied by detailed drawings of pipe interface positioning. Project-level rule layer: Supplement special clauses based on the project's individual needs, such as adding a transportation direction diagram for oversized components, and marking the seismic grade on the detailed drawings of nodes in areas with a seismic fortification intensity of 8 degrees. All rules are coded, and the component types and drawing types to which the rules apply are labeled to form a rule list; drawing types include layout drawings, detail drawings, production drawings, and acceptance drawings; S202. Rule Parametric Conversion: Convert the rule list into a machine-readable parametric format. Develop a rule parsing engine through the BIM platform's open API interface. The specific implementation process is as follows: The text specifications are converted into structured rule expressions using Python scripts. For example, the 1:2 scale detail of the horizontal joint node of the precast wall panel is converted into a logical judgment condition: when the component type is a precast wall panel and the connection type is a horizontal joint, the view scale is forcibly set to 1:2. The rule triggering matrix is ​​constructed based on two dimensions: component type and drawing type. The matrix row dimension is divided into four categories based on component type: precast concrete components, steel structure components, electromechanical integrated components, and decorative integrated components. The column dimension is set into four categories based on drawing type: layout drawing, detail drawing, production drawing, and acceptance drawing. The corresponding parametric expressions of the rules are stored in the matrix cells. National-level mandatory rules are marked in red and given the highest level of enforcement authority; enterprise-level rules are marked in yellow and allow project-level adjustment of parameter thresholds; project-level rules are marked in blue and granted temporary disabling authority. A rule execution engine was developed based on the open API interface of the BIM platform. The engine has a built-in rule trigger matrix and a two-way communication channel for component parameters. When the user initiates a drawing command, the engine automatically obtains the type, attributes and drawing type of the target component, and triggers the corresponding rule set through matrix matching. The rule judgment is processed in parallel based on multi-threading technology. For example, when generating a layout drawing for precast concrete wall panel components, national-level rules, enterprise-level rules and project-level rules are triggered simultaneously. The execution process implements hierarchical access control. Among them, national-level rules marked in red cannot be skipped and automatically override conflicting rules, enterprise-level rules marked in yellow allow users to adjust parameter thresholds through the plugin interface, and project-level rules marked in blue provide a temporary disabling option. The results of rule execution are fed back through the BIM plugin's visual interface. Component drawing requests that fail to comply with the mandatory rules will be blocked and a rectification list will pop up. At the same time, the rule execution log is recorded in the central database. S203, Rule Base Maintenance and Optimization Mechanism: The performance evaluation of rule usage is based on the rule execution log of the BIM platform, using a formula. The rule applicability index is calculated. ,in, Indicates the number of times the rule is triggered. Total number of plots The ratio, i.e., the rule trigger frequency, Indicates the first Secondary user adjustment consistency refers to the consistency between the initial output map automatically generated by the rule engine and the final output map after manual adjustment by the user; This indicates the total number of times the user manually adjusted the rules after they were triggered. Indicates the first The number of times a rule conflicts with other rules. This represents the total number of conflict events that occur between all rules. Do not assign weighting factors to rule triggering frequency, user adjustment consistency, and rule conflict rate; When the rule applicability index is lower than the preset threshold, the current rule is marked as needing optimization, and the rule administrator is notified to analyze it. If the rule applicability index is lower than the preset threshold multiple times and optimization is ineffective, the rule will be directly disabled or eliminated. All problematic rules are ranked based on the rule applicability index, and a rule base health report is generated, which includes rule applicability ranking, conflict rate ranking, a list of rules to be optimized, and optimization suggestions.

[0021] S3. Automated Generation and Association of Drawings: Based on fully verified component information and parametric drawing rules, standard drawings are automatically generated through the BIM platform, and a two-way association relationship is established between the drawings, components, and the BIM platform. The specific implementation process is as follows: S301, Intelligent Analysis and View Generation for Drawing Output Tasks: Receive the drawing instructions processed by the S2 step rule engine, parse the target component set and the corresponding drawing type requirements; determine the drawing scale, viewport range and section position based on the component spatial positioning data and view rules. For layout drawings, a spatial grid algorithm is used to automatically arrange plan and elevation views, and the display priority is controlled based on component functional hierarchy, generation process and installation stage; For detailed drawings, node construction parameters are extracted based on rule-triggered matrices to automatically generate sectional views and enlarged partial views; For production drawings, the system retrieves component manufacturing process data to generate detailed fabrication drawings, including pouring joints, rebar arrangement, and embedded part positioning. A dynamic priority system based on rule weights and component production urgency is constructed using a formula. The drawing priority index is calculated. ,in, This indicates the inherent weight of the rule: 1.0 for national-level rules, 0.7 for enterprise-level rules, and 0.5 for project-level rules. This indicates the urgency of the situation. These are the component production deadline and the current time, respectively. To ensure the completeness of component information, These are the influence weight factors of the inherent weight of the rule, the time urgency, and the completeness of the component information, respectively; Based on the principle of sorting the priority index of drawings from high to low, drawings with higher priority index are generated first; after the drawings are generated, they are automatically associated with BIM platform data based on component ID, and clicking on the component in the drawing supports highlighting the corresponding entity in the BIM platform. S302, Parametric Annotation Driven: Dimensioning layer: Based on the 3D dimensional accuracy of components, hole coordinates and tolerance data extracted from the geometric feature layer, dimension chains and tolerance annotations are generated; when the component is an over-limit component, the transport direction annotation rule is triggered; Technical annotation layer: Calls data from the material performance layer and connection characteristic layer to annotate concrete strength grade, weld size, and sealing material performance; for nodes in seismic fortification areas, it supplements the annotation of seismic structural requirements; Symbol labeling layer: Based on the enterprise standard library, a predefined symbol set is embedded, including lifting point marks, installation direction arrows and weld symbols. The symbol positions are dynamically adjusted based on the component geometric center intelligent avoidance algorithm. The annotation results are synchronized with the BIM platform data in real time through hash verification. When the BIM platform parameters change, the annotations are automatically updated and version differences are recorded. S303, Multi-disciplinary Drawing Collaboration and Association: Establish a three-dimensional association database of drawings, components, and platforms, and use lightweight technology to achieve two-way linkage between drawings and the BIM platform. Support direct viewing of the dynamic attributes of associated components in the drawing interface, including hoisting sequence and installation coordinates. Based on the cloud platform, the system enables the synchronous generation and conflict detection of architectural, structural, and mechanical and electrical (M&E) drawings. For example, when there is a conflict between the embedded M&E pipelines and the structural reinforcement, a cross-disciplinary adjustment warning is automatically triggered and the associated drawings are modified synchronously. A drawing version ID is generated and bound to the component version number to ensure the consistency between the drawing and the version. Each drawing output operation automatically generates a version log, recording the output time, the triggered rule set, and the associated component ID. Based on the project's drawing standards, the drawing frame, signature column, and drawing catalog are assembled, and the drawing frame information is automatically filled based on the component attributes. It supports parallel output of multiple formats, automatically divides the drawing packages based on the drawing rules, and uses blockchain technology to digitally sign the drawing packages to ensure tamper-proof protection.

[0022] S4. Intelligent verification and dynamic optimization of drawing quality: Based on intelligent algorithms and rule engines, multi-dimensional quality verification is performed on generated drawings, and dynamic optimization and feedback learning are carried out on the problems found. The specific implementation process is as follows: Based on the parameterized drawing rule library built on S2, a mandatory rule compliance scan is performed on the generated drawings; image recognition technology is used to parse the drawing elements and detect standardization issues, including missing annotations, incorrect scale, and misuse of symbols. Convolutional neural networks are used to extract features from drawings and views, and to identify the semantic consistency of annotation text and symbols, such as detecting whether the deviation between dimension annotations and actual component dimensions exceeds the tolerance range. The technical description text is parsed using natural language processing algorithms, and the matching degree between the technical description text and the component material performance layer and connection characteristic layer data is verified. For example, when the marked concrete strength grade is inconsistent with the parameters stored on the BIM platform, an orange warning is triggered and the data source difference is indicated. Generative adversarial networks are introduced to simulate the logic of manual image review, generating virtual error samples to train the verification algorithm and improve the sensitivity to detect hidden problems, including line confusion and legend overlap. The verification algorithm interacts with the drawing data in real time through the open API interface of the BIM platform. Each drawing generates a unique verification ID, which is then bound to the component ID. Based on the detection results, a rule-based hierarchical drawing error classification, early warning, and handling mechanism is constructed: Level 1 Warning: If a violation of mandatory national regulations affects structural safety, immediately halt the drawing release process, trigger an audible and visual alarm, and notify the project manager and relevant designers through the message center to rectify the situation immediately. Level 2 warning: Issues that violate enterprise-level standard rules will generate a list of suggested modifications, including the location of the issue, the rule clause, and the suggested modification value. Designers are allowed to make adjustments within the allowable threshold range based on the actual situation, but the reasons for the adjustments must be recorded. Level 3 warning: Violation of project-level special rules. A prompt and optimization suggestion will be provided. Designers can choose to adopt or temporarily disable the optimization suggestion, and the choice will be recorded for subsequent rule effectiveness evaluation. The verification results are visualized through a dedicated quality inspection panel in the BIM plugin. The panel displays all issues in a list format, with each issue accompanied by a description, rule basis, severity level, and location button. After clicking the location button, the problematic element is automatically highlighted in the drawing viewport, and can be further traced back to the corresponding entity in the BIM platform through the component ID, enabling rapid linkage and two-way modification between drawing issues and platform information. All verification operations and results are recorded and written to the central database. A feedback dataset is built based on the verification logs to record each type of problem, optimization measures, and final adoption results. Based on a preset cycle, rule performance indicators are collected periodically, including rule false positive rate, false negative rate, and user adoption rate. When the rule false positive rate continues to be higher than the preset threshold, the weight of the current rule in the verification process is automatically reduced, and the rule administrator is notified to review it. The first-time pass rate of drawings serves as a reward signal, and the rule trigger threshold is dynamically adjusted; through formulas... The reward coefficient for a single map output task is calculated. ,in, This indicates the number of serious errors generated in this drawing production task, that is, the number of violations of mandatory national rules; This indicates the number of general errors generated during this drawing output task, i.e., the number of violations of enterprise-level or project-level rules; When the reward coefficient When =10, it means that the drawing passes all checks at once and is set as the highest optimization target; Through formula The cumulative discount reward coefficient is calculated, where, Indicates the first One drawing output task; Indicates the first The discount factor for the reward coefficient of each map-producing task; The optimal strategy is obtained based on the cumulative discount coefficient. The rule triggering threshold is then adjusted to the best parameters based on the optimal strategy, thereby maximizing long-term rewards in complex project environments, i.e., achieving the highest first-time pass rate for drawings. A rule optimization report is generated quarterly, listing the rules to be adjusted and providing an impact analysis, using formulas... The comprehensive index affecting the calculation is obtained, among which, This represents the set of component types constrained by the current rule. This represents the weight of component type b in the project. This represents the average drawing frequency of component type b; The higher the comprehensive impact index, the wider the scope of the current rule's influence. It is used to predict the impact of a rule change on the project before a drawing rule is modified or optimized. At the same time, drawing rules with high comprehensive impact indices are prioritized based on the comprehensive impact index. The optimization report initiates the revision process directly through the collaborative interface of the BIM platform. Relevant experts or technical personnel will receive notifications on the same interface and can conduct online reviews, comments, and votes on the proposed rule modifications. The revision process is fully recorded, and the new rule version must undergo simulation testing and authorization approval before it takes effect. Based on the verification results, the distribution of various errors is statistically analyzed, and a quality score radar chart is generated based on the classification of architecture, structure, and electromechanical disciplines. For blocking errors, create a collaborative task sheet, assign it to the relevant designers, track the rectification progress, and escalate the reminder to the project manager if it is not handled within the time limit. The cloud platform enables multi-disciplinary quality inspection data sharing. When optimization of professional drawings causes conflicts between related disciplines, a cross-disciplinary verification session is automatically triggered, and the collaborative solution is recorded. A drawing quality certification label is generated, and only drawings that pass all mandatory verifications can be published. The label information and the drawing version ID are stored together on the blockchain.

[0023] This invention relates to an intelligent drawing system for prefabricated buildings based on BIM technology, comprising: Component Information Processing Module: This module acquires a comprehensive component family library from the BIM platform, including precast concrete, steel structure, electromechanical integration, and decorative integrated components. Each component is labeled with a unique ID, and production, installation, and operation and maintenance parameters are stored synchronously. Geometric features, material properties, and connection characteristics are extracted layer by layer via the BIM open API. Dynamic attributes of components are extracted in real-time using Python scripts, and a structured dataset is generated using JSON with the component ID as the key. A dynamic mapping matrix between components and drawing elements is constructed, and combined with a four-level verification chain, a completion list and a completeness report are generated. The rules management module constructs a three-tiered rule framework: national level extracts mandatory requirements from industry standard databases, enterprise level refines production clauses based on enterprise standards, and project level supplements special requirements. All rules are coded and labeled with their applicable scope to form a list. It uses Python scripts to convert text rules into parameterized expressions and constructs a rule trigger matrix based on component type and drawing type. It sets hierarchical access permissions and develops a rule execution engine with a built-in two-way communication channel. Based on execution logs, it calculates a rule applicability index and generates a rule base health report including rule rankings and a list of rules to be optimized. Drawing generation and association module: Parses drawing instructions, determines drawing scale and viewport range based on component spatial positioning; Layout drawings arrange plan and elevation views based on spatial grid algorithm, and control display priority based on functional hierarchy; Detail drawings extract node parameters to generate sectional views; Production drawings call process data to generate processing drawings including pouring joints and rebar layout; The drawing order is sorted based on drawing priority index, parameterized driving annotation, and a 3D association database is built to achieve two-way linkage between drawings and BIM; The cloud platform detects multi-disciplinary conflicts, outputs multi-format drawings, and uses blockchain signature to prevent tampering; Quality Verification and Optimization Module: Verifies drawings using multiple algorithms: scanning mandatory clauses for rule compliance, detecting missing annotations using image recognition, identifying dimensional deviations using convolutional neural networks, parsing technical descriptions of BIM parameter matching using natural language processing, and identifying hidden problems using adversarial networks; Establishes a tiered early warning system: Level 1 (national-level violations) blocks publication and triggers audio-visual alarms; Level 2 (enterprise-level violations) generates modification suggestions; Level 3 (project-level violations) provides optimization solutions; Calculates cumulative discount reward coefficients based on verification logs to adjust rule thresholds, and generates quarterly optimization reports; Generates professional quality radar charts, tracks and rectifys blocking errors, and generates certification labels for only drawings that pass all mandatory verifications, storing them along with the version ID on the blockchain.

[0024] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for intelligent drawing generation of prefabricated buildings based on BIM technology, characterized in that, include: S1. Structured extraction and mapping of component information: In the BIM platform environment, the multi-source information of the prefabricated building platform is extracted in a structured manner, and a mapping relationship with the drawing elements is established. S2. Parametric construction of drawing rules: Based on the design specifications for prefabricated buildings and enterprise standards, a parametric drawing rule library is constructed in the BIM supporting plugin; S3. Automated generation and association of drawings: Based on the verified component information and parametric drawing rules, standard drawings are automatically generated through the BIM platform, and a two-way association relationship is established between the drawings, components and the BIM platform. S4. Intelligent verification and dynamic optimization of drawing quality: Based on intelligent algorithms and rule engines, multi-dimensional quality verification of generated drawings is performed, and dynamic optimization and feedback learning are carried out on the problems found.

2. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific implementation process of step S1 is as follows: Obtain the full-discipline component family library of the prefabricated building BIM platform, including precast concrete, steel structure, electromechanical integration and decoration integrated components, and each component is marked with a unique ID; the full-discipline component family library also includes parameters for production, installation and operation and maintenance in all dimensions; By establishing real-time communication with the digital twin of the component through the open API interface of the BIM platform, information on the geometric feature layer, material performance layer, and connection characteristic layer is extracted layer by layer. Dynamic attributes of the component are extracted based on Python scripts and stored in JSON format. A structured dataset is constructed with the component ID as the unique key value. Based on the four dimensions of function, production, installation and operation and maintenance, component information is sorted out and a dynamic adaptation mapping matrix between component parameters and drawing elements is constructed. A four-level verification chain is constructed: Level 1 uses a hash algorithm for real-time verification; Level 2 uses a knowledge graph for correlation verification; Level 3 uses a neural network combined with an industry standard database for integrity verification; and Level 4 calls the industry standard database for compliance verification. After verification, a supplementary list and a component information completeness report are generated.

3. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific operation process of step S2 includes: Construct a three-tiered rule framework encompassing specifications, standards, and projects: National-level rules extract mandatory requirements from industry standard databases, including detailed proportional drawings of horizontal joint nodes in precast wall panels and marking of weld leg dimensions and lengths for steel structure welds; enterprise-level rules refine production adaptability clauses based on enterprise standards; project-level rules supplement clauses based on project-specific needs. Code all rules, label the applicable component types and drawing types, and form a rule list; The textual specifications of the rule list are converted into structured rule expressions using Python scripts; a rule trigger matrix is ​​constructed based on two dimensions: component type and drawing type, and the matrix cells store the corresponding parameterized expressions of the rules. National-level rules are marked in red to set the highest mandatory enforcement authority, enterprise-level rules are marked in yellow to allow adjustment of parameter thresholds, and project-level rules are marked in blue to enable temporary disabling authority; a rule execution engine is developed, which has a built-in rule trigger matrix and a two-way communication channel for component parameters. When a user initiates a drawing command, the rule set is automatically matched. During execution, it is controlled based on hierarchical permissions. Drawing requests that do not pass the mandatory rules are blocked and a rectification list is displayed. At the same time, the rule execution log is recorded to the central database.

4. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific operation process of step S2 also includes: The performance evaluation of rule usage is based on the rule execution log of the BIM platform, using a formula. Obtain the rule applicability index ,in, Indicates the number of times the rule is triggered. Total number of plots The ratio, i.e., the rule trigger frequency, Indicates the first Secondary users adjust the fit; This indicates the total number of times the user manually adjusted the rules after they were triggered. Indicates the first The number of times a rule conflicts with other rules. This represents the total number of conflict events that occur between all rules. Do not assign weighting factors to rule triggering frequency, user adjustment consistency, and rule conflict rate; When the rule applicability index is lower than the preset threshold, the current rule is marked as needing optimization; if it is lower than the preset threshold multiple times and optimization is ineffective, the rule is directly disabled or eliminated. Generate a rule base health report, including rule applicability ranking, conflict rate ranking, a list of rules to be optimized, and optimization suggestions.

5. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific implementation process of step S3 is as follows: Receive the drawing instructions processed by the S2 step rule engine, identify the target component set and drawing type requirements, and determine the drawing scale, viewport range and section position. Layout drawings use a spatial grid algorithm to arrange plan and elevation views and control display priority; detail drawings extract node construction parameters to generate sectional views and enlarged partial views; production drawings call production process data to generate processing drawings. Through formula Obtain drawing priority index ,in, This indicates the inherent weight of the rule; This indicates the urgency of the situation. These are the component production deadline and the current time, respectively. To ensure the completeness of component information, These are the influence weight factors of the inherent weight of the rule, the time urgency, and the completeness of the component information, respectively; The dimensioning layer generates dimension chains and tolerance annotations based on geometric feature data, and oversized components trigger transport direction annotations; the technical annotation layer calls material and connection data to annotate performance parameters and seismic requirements; the symbol annotation layer embeds predefined symbols and adjusts their positions; BIM data is synchronized through hash verification, and annotations are automatically updated when parameters change; Establish a 3D database linking drawings, components, and the platform, and achieve two-way linkage based on lightweight technology; the cloud platform synchronously generates multi-disciplinary drawings and detects conflicts, generates drawing version IDs to bind component versions, and records version logs.

6. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific implementation process of step S4 includes: Based on the parameterized drawing rule library built on S2, the mandatory rule compliance of the generated drawings is scanned; image recognition is used to detect standardization issues, view features are extracted based on convolutional neural networks to verify dimensional deviations, technical description text is parsed based on natural language processing algorithms, and adversarial network training algorithms are introduced to detect line type confusion and legend overlap; the verification algorithm interacts with the drawing data in real time through the BIM platform API interface, and a unique verification ID is generated for each drawing and bound to the component ID; A three-tiered early warning mechanism is established: Level 1 warning blocks non-compliant drawings that could affect structural safety and triggers audible and visual alarms; Level 2 warning generates enterprise-level rule modification suggestions; and Level 3 warning provides project-level rule optimization suggestions. A feedback dataset is constructed based on the verification logs, using the formula... The reward coefficient for a single map output task ,in, This indicates the number of critical errors that occurred during this drawing output task; This indicates the number of general errors generated during this drawing output task.

7. The intelligent drawing generation method for prefabricated buildings based on BIM technology according to claim 1, characterized in that, The specific implementation process of step S4 also includes: Through formula The cumulative discount reward coefficient is obtained, where, Indicates the first One drawing output task; Indicates the first The discount factor for the reward coefficient of each map-producing task; The optimal strategy is derived based on the cumulative discount reward coefficient; a rule optimization report is generated quarterly, using a formula. The comprehensive index of influence is obtained, among which, This represents the set of component types constrained by the current rule. This represents the weight of component type b in the project. This represents the average drawing frequency of component type b; Based on the verification results, various error distributions are statistically analyzed to generate a professional quality radar chart. The cloud platform shares data and triggers cross-professional verification. The drawings that are subject to mandatory verification generate quality certification labels, which are stored on the blockchain along with the version ID.

8. A system applied to the intelligent drawing generation method for prefabricated buildings based on BIM technology as described in any one of claims 1-7, comprising: Component Information Processing Module: Obtains the full-discipline component family library from the BIM platform, generates a structured dataset, extracts geometric, material, and connection information layer by layer through API interfaces, obtains dynamic attributes based on Python scripts, and stores them in JSON format; Construct a dynamic mapping matrix between components and drawing elements, incorporate a four-level verification chain, and generate a completion list and a completeness report; Rule management module: Constructs a three-tier rule framework of specifications, standards, and projects; converts rules into parameterized expressions based on Python scripts; and builds a trigger matrix. Set up permission levels, develop a rule execution engine, and generate health reports; Drawing generation and association module: Parses drawing instructions to generate views, and arranges the drawing order based on the drawing priority index; Parametrically driven dimensions, techniques, and symbol annotations; Build a 3D relational database, output multi-format drawings, and perform blockchain signature; Quality verification and optimization module: Verifies the compliance of drawings based on multiple algorithms and constructs a hierarchical early warning mechanism; Optimize rules based on log feedback, and adjust thresholds based on cumulative discount reward coefficients; Generate quality reports and certification labels.

Images