A process-level multidimensional coding construction document inspection system and method

Abstract

This invention relates to the field of information technology in engineering construction, and in particular to a process-level multi-dimensional coded construction document inspection system and method. The system includes a four-dimensional dynamic coding engine, an intelligent document verification module, a blockchain evidence storage and repair module, and a dynamic traceability platform. The four-dimensional dynamic coding engine generates unique codes and verifies process logic. The intelligent document verification module identifies missing documents based on an engineering document template library and a weighted algorithm. The blockchain evidence storage and repair module stores missing records on the blockchain and sets a time-limited completion mechanism. The dynamic traceability platform realizes bidirectional mapping and visual monitoring between construction documents and BIM models. The method, through the collaboration of these modules, completes code generation verification, document integrity detection, missing evidence storage and repair, and traceability of the entire construction process. This invention solves the problems of low efficiency in traditional construction document management, difficulty in process supervision, and difficulties in data traceability and collaboration, promoting the digital transformation of engineering management and improving engineering quality and management efficiency.

Description

Technical Field

[0001] This invention relates to the field of information technology in engineering construction, specifically to a process-level multi-dimensional coded construction document detection system and method. Background Technology

[0002] In the context of the current rapid development of the engineering construction industry, the standardization and accuracy of construction document management are crucial for project quality, schedule control, and subsequent acceptance auditing. However, traditional construction document management methods have many problems that urgently need to be addressed, becoming a bottleneck restricting the efficient development of the industry. These problems are specifically reflected in the following aspects:

[0003] Document management is inefficient and prone to errors: In traditional construction processes, the management of construction documents relies heavily on manual operations and paper records. From document collection, sorting to archiving, it not only consumes a lot of manpower and time, but also easily leads to problems such as missing or incorrect documents. In large and complex engineering projects, due to the involvement of numerous construction procedures and participants, the number of documents is huge and the types are diverse. Manual management is difficult to ensure the completeness and accuracy of documents, which leads to a lot of time being spent on document checking, filling in and correcting omissions during subsequent project acceptance and auditing, seriously affecting project progress and efficiency.

[0004] Lack of effective means to supervise the logical relationship of construction processes: The reasonable arrangement and execution of construction processes are the key to ensuring project quality and safety; however, the existing construction management model lacks effective means to supervise the logical relationship of processes, making it difficult to judge in real time whether the processes are carried out in the standard order. This can easily lead to situations such as reversed processes and incorrect parallel processes, which can cause potential quality hazards and safety accidents. Once a problem occurs, it is difficult to trace back to the source of the problem and take effective measures to rectify it.

[0005] Data traceability and collaboration challenges: As the scale of engineering construction projects continues to expand and the number of participants increases, the frequent information exchange between these participants places higher demands on the traceability and collaborative management of construction data. In the traditional approach, construction data is stored in a scattered manner, lacking a unified management and traceability mechanism, resulting in poor information sharing and low collaboration efficiency among the participants. When quality problems or disputes arise in the project, it is difficult to quickly and accurately trace relevant construction documents and process information, affecting the efficiency of problem-solving and the determination of responsibility.

[0006] Lagging technology application and outdated management models: Although digital technology has been applied in the field of engineering construction, existing construction document management systems are mostly single-function modules, lacking the ability to integrate and collaboratively manage information from the entire construction process and multiple dimensions; in terms of coding technology, existing coding methods cannot fully cover the key information of construction procedures, making it difficult to achieve accurate association between construction documents and engineering entities; in terms of data security and evidence preservation, traditional storage methods cannot guarantee the authenticity, integrity, and immutability of data, failing to meet the high requirements of engineering construction for data management.

[0007] Therefore, a process-level multidimensional coding construction document detection system and method are proposed to address the above problems. Summary of the Invention

[0008] The purpose of this invention is to provide a process-level multidimensional coded construction document detection system and method to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A process-level multidimensional coding construction document inspection system includes:

[0011] The four-dimensional dynamic coding engine is used to generate unique coding identifiers consisting of engineering type code, structural part code, process sequence code, and spatiotemporal coordinate code, and to verify the legality of process logical relationships through a topological constraint verification algorithm.

[0012] Intelligent file verification module: Generates a file integrity matrix based on a pre-set project template library, uses an improved Levenshtein distance algorithm to match missing files and outputs a replacement list;

[0013] Blockchain Evidence Preservation and Repair Module: Preserves missing file records on the blockchain and triggers a timed completion mechanism;

[0014] Dynamic traceability platform: It links the coding engine and the verification module to achieve visualized monitoring of the entire lifecycle of construction documents.

[0015] As a preferred approach, the topology constraint verification algorithm is based on an improved directed acyclic graph model, and its verification process is as follows:

[0016] Calculate the sum of the products of the time difference function values ​​of all adjacent process nodes in the process chain to be verified and the process weight factor. When the sum is greater than or equal to the validity threshold, output a verification pass signal; otherwise, output a verification failure signal.

[0017] The time difference function outputs a value of 1 when the timestamp of the subsequent process is greater than the timestamp of the preceding process, and outputs a value of 0 otherwise.

[0018] Process weighting factors are assigned values ​​according to the importance of each process.

[0019] The validity threshold is equal to the total number of nodes in the process chain.

[0020] As a preferred approach, the intelligent file verification module executes the weighted Levenshtein distance algorithm:

[0021] Calculate the minimum difference value between the actual file set and each project template. This difference value consists of two parts:

[0022] The first part is the sum of the absolute value of the difference between the actual file status and the template file status multiplied by the file importance weight;

[0023] The second part is the product of the hierarchy decay factor and the hierarchy depth of the missing file in the project structure tree;

[0024] The document importance weight is assigned a graded value based on the document's criticality, and the level depth corresponds to different values ​​for unit project, sub-item project, and process level.

[0025] As a preferred approach, the construction method for the engineering template library includes:

[0026] Analyze the component features of historical engineering drawings, decompose them according to the four-level structure of unit project, sub-project, sub-item project and process category, and generate a file tree corresponding to the coding level;

[0027] By linking the construction specification table with the process sequence code, a dynamically updatable template knowledge graph is formed.

[0028] As a preferred solution, the blockchain evidence storage and repair module performs the following operations:

[0029] A lightweight Merkle tree verification algorithm is adopted, with a verification path complexity of logarithmic order.

[0030] The evidence storage data package includes the hash value of the previous block, a four-dimensional encoded identifier, the type of missing file, a timestamp, and the ID of the person responsible.

[0031] Data validation employs a two-factor authentication mechanism:

[0032] The first factor uses the Ed25519 algorithm to verify the signature of the combination of the encoded identifier and the timestamp;

[0033] The second factor uses the national cryptographic algorithm SM2 to verify the identity of the responsible person through signature verification;

[0034] Both verification results must pass for the data to be considered valid.

[0035] As a preferred solution, a dynamic traceability platform includes:

[0036] Establish a two-way mapping relationship between two-dimensional construction drawings and BIM models, and use spatiotemporal coordinate codes as indexes to locate the component process progress;

[0037] An improved YOLOv5s model is used to identify coded regions in drawings, and its loss function consists of three parts.

[0038] The classification loss uses the focus loss function to address the class imbalance problem;

[0039] The target detection loss is multiplied by a coefficient of 0.8;

[0040] The encoding recognition enhancement loss is multiplied by a factor of 1.5, and this enhancement loss is specifically designed for the four-dimensional encoded character region.

[0041] As a preferred option, the assignment rule for the process weight factor is as follows:

[0042] Basic process types include template installation and measurement and layout, which are assigned benchmark weight values;

[0043] Key quality control points include concrete pouring and prestressing tensioning, which are given a weight twice that of the baseline value.

[0044] Safety control procedures include foundation pit support and blasting operations, which are assigned a weight 2.5 times that of the baseline value.

[0045] As a preferred solution, the time-limited completion mechanism includes a dynamic early warning strategy:

[0046] When a missing file is detected for the first time, a warning notification is pushed to the mobile terminal of the person in charge.

[0047] If not processed within 24 hours, the project chief engineer's terminal will be notified simultaneously.

[0048] If not handled within forty hours, an alarm at the project management system level will be triggered;

[0049] If the issue remains unresolved for more than 48 hours, an unalterable traceability certificate will be generated and the relevant acceptance process will be locked.

[0050] As a preferred approach, bidirectional mapping is achieved through spatial coordinate transformation:

[0051] The spatiotemporal coordinate code is decomposed into two parts: the construction date code and the geographical coordinate abbreviation code;

[0052] The method for mapping geographic coordinate abbreviations to BIM model spatial grids is as follows:

[0053] Calculate the difference between the actual longitude and the longitude of the project origin, divide by the grid side length, and round down.

[0054] Calculate the difference between the actual latitude and the project origin latitude, divide by the grid side length, and round down.

[0055] Multiply the result calculated in the longitude direction by 1000 and add it to the result calculated in the latitude direction to obtain the unique identifier of the spatial grid.

[0056] A method for detecting multi-dimensional coded construction documents at the process level is provided. This method utilizes a multi-dimensional coded construction document detection system at the process level to detect multi-dimensional coded construction documents at the process level.

[0057] As can be seen from the above technical solutions provided by the present invention, the beneficial effects of the process-level multidimensional coded construction document detection system and method provided by the present invention are:

[0058] Precision construction management: The four-dimensional dynamic coding engine assigns a unique code identifier to each construction process, covering information such as project type, structural part, process sequence and spatiotemporal coordinates, and verifies the logical relationship of the process by combining topological constraint verification algorithm; This enables construction managers to clearly grasp the construction progress, accurately locate process problems, effectively avoid situations such as process reversal and parallel errors, ensure that construction is carried out strictly in accordance with the predetermined plan and specifications, and significantly improve the level of precision and scientific management of construction.

[0059] Highly efficient document integrity assurance: The intelligent document verification module, based on the engineering template library and the weighted Levenshtein distance algorithm, accurately identifies missing documents and generates a replacement list; compared with traditional manual inspection, it greatly improves the efficiency of document inspection, reduces labor costs, and ensures the completeness and standardization of construction documents, providing a solid guarantee for project acceptance, auditing and other work, and effectively avoiding compliance risks caused by missing or non-standard documents;

[0060] Trusted data storage and closed-loop problem handling: The blockchain storage and repair module uses blockchain technology to store missing file records. It ensures that the data is tamper-proof and fully traceable through a lightweight Merkle tree verification algorithm and two-factor authentication. The time-limited completion mechanism and dynamic early warning strategy form a closed loop for problem handling, from the initial push to the responsible person to locking the acceptance process after the timeout. This strengthens the implementation of responsibilities, ensures the compliance of project document management, and effectively prevents data tampering and management loopholes.

[0061] Full-process visual dynamic traceability: The dynamic traceability platform achieves precise association between construction documents, process progress and engineering entities through BIM-coding bidirectional mapping and improved YOLOv5s model, supporting visual monitoring and historical data backtracking of the entire construction process; engineering managers can intuitively view construction progress and document status, quickly locate problem nodes, provide strong data support for engineering quality traceability, review and optimization, and significantly improve the transparency and collaborative efficiency of engineering management.

[0062] Driving the digital transformation of engineering: This invention integrates advanced technologies such as multidimensional coding, blockchain, BIM, and computer vision to build an intelligent and digital construction document inspection system. Through data interaction and collaborative work between modules, it breaks down information silos, promotes collaborative management among construction units, construction companies, and supervision units, and provides core technical support for the engineering industry to transform from traditional management models to digital and intelligent ones, thus contributing to the high-quality development of the industry. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the overall structure of a process-level multidimensional coded construction document detection system according to the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0065] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific embodiments.

[0066] like Figure 1 As shown, this embodiment of the invention provides a process-level multidimensional coded construction document detection system and method, including:

[0067] The four-dimensional dynamic coding engine is used to generate unique coding identifiers consisting of engineering type code, structural part code, process sequence code, and spatiotemporal coordinate code, and to verify the legality of process logical relationships through a topological constraint verification algorithm.

[0068] Intelligent file verification module: Generates a file integrity matrix based on a pre-set project template library, uses an improved Levenshtein distance algorithm to match missing files and outputs a replacement list;

[0069] Blockchain Evidence Preservation and Repair Module: Preserves missing file records on the blockchain and triggers a timed completion mechanism;

[0070] Dynamic traceability platform: It links the coding engine and the verification module to achieve visualized monitoring of the entire lifecycle of construction documents.

[0071] In this embodiment, the topology constraint verification algorithm is based on an improved directed acyclic graph model, and its verification process is as follows:

[0072] Calculate the sum of the products of the time difference function values ​​of all adjacent process nodes in the process chain to be verified and the process weight factor. When the sum is greater than or equal to the validity threshold, output a verification pass signal; otherwise, output a verification failure signal.

[0073] The time difference function outputs a value of 1 when the timestamp of the subsequent process is greater than the timestamp of the preceding process, and outputs a value of 0 otherwise.

[0074] Process weighting factors are assigned values ​​according to the importance of each process.

[0075] The validity threshold is equal to the total number of nodes in the process chain;

[0076] Furthermore, the four-dimensional dynamic coding engine, as a core component of the process-level multi-dimensional coding construction document inspection system, bears the important responsibility of generating unique coding identifiers and verifying the legality of process logic. It is crucial for ensuring the efficient operation of the system and achieving accurate traceability. The following will elaborate on its overall function, components, key technical principles, workflow, and application value:

[0077] I. Overall Function Overview:

[0078] The core function of the four-dimensional dynamic coding engine is to generate a unique code identifier for each process in the construction process, containing information on the project type, structural part, process sequence, and spatiotemporal coordinates, which serves as the "digital ID card" for the construction documents. At the same time, it uses a topological constraint verification algorithm to verify the logical relationships between processes, ensuring that the construction process meets the specifications. In addition, the codes generated by the coding engine can also be used as indexes to associate construction documents with project entities, providing basic data support for the full lifecycle management of the construction process.

[0079] II. Submodule Composition and Functions:

[0080] (a) Multidimensional coding generation unit:

[0081] Encoding rule definition: According to the established encoding structure, construction information is broken down into project type code (2 digits), structural part code (4 digits), process sequence code (3 digits), and spatiotemporal coordinate code (6 digits). The project type code is used to distinguish different types of projects, such as "01" representing building construction and "02" representing highway construction. The structural part code accurately identifies the specific location in the project, such as "0301" representing the main structure - concrete engineering. The process sequence code is numbered sequentially according to the construction order, such as "001" representing formwork installation. In the spatiotemporal coordinate code, the first 3 digits represent the construction date (such as "225" representing the 5th day of 2022), and the last 3 digits are the GPS coordinate abbreviation.

[0082] Dynamic code generation: Based on the actual information of each process during construction, a unique code identifier is generated in real time; at the beginning of a process, the system automatically collects information such as project type, structural part, construction time and spatial location, combines and generates a code according to the coding rules, and binds it to the construction documents related to that process;

[0083] (ii) Intelligent document verification unit:

[0084] Data Acquisition and Processing: Collect timestamp information and process attributes (distinguishing between critical processes, ordinary processes, safety control processes, etc.) for each process to provide a data foundation for topology constraint verification; organize this data into a process chain as input for the verification algorithm.

[0085] Topology constraint verification algorithm execution: A verification formula based on an improved directed acyclic graph (DAG) model is used. (in, The process chain to be verified; For the first Each process node timestamp; Let the time difference function between adjacent processes be used. Output 1 if the condition is met, otherwise output 0. The process weighting factor is set to 1 for basic processes, 2 for critical quality control point processes, and 2.5 for safety control processes. The validity threshold is set to the length of the process chain. Perform validation of the logical relationships between work processes; for example, for a process containing 3 key work processes... The process chain, threshold The process chain is considered valid only when the time difference between adjacent processes is positive and the sum of their weights is greater than or equal to 3.

[0086] Verification result feedback: If the verification result is valid ( If the verification result is invalid, the corresponding construction documents in that process chain are allowed to proceed to the next processing step; otherwise, the verification result is invalid. If the error occurs, an early warning message will be issued immediately, indicating to the construction personnel that there is a problem with the logical process of the construction procedure, which needs to be adjusted, and the circulation of relevant construction documents will be stopped.

[0087] (III) Encoding Management and Storage Unit:

[0088] Encoding Storage: The generated unique encoding identifier and its corresponding process information, construction document index, and other data are stored in the system's database; an efficient data storage structure is adopted to ensure fast querying and retrieval of encoding data;

[0089] Coding update and maintenance: When design changes or work procedures are adjusted during construction, the corresponding codes should be updated and maintained in a timely manner; the updated coding information should be synchronized to the database and associated construction documents to ensure data consistency and accuracy; at the same time, the code change history should be recorded for subsequent traceability and auditing.

[0090] III. Key Technology Principles:

[0091] (I) Principles of Multidimensional Coding Technology:

[0092] Multidimensional coding technology digitizes key information in the engineering construction process. Through specific coding rules, information such as project type, structural parts, process sequence, and spatiotemporal coordinates are compressed into a single code. This coding method is unique and systematic, accurately identifying each construction process and enabling rapid location and retrieval of construction information. At the same time, the multidimensional structure of the code allows information from different dimensions to be interconnected, facilitating comprehensive management of the construction process.

[0093] (II) Principles of Topology Constraint Verification Technology:

[0094] Topology constraint verification technology is based on a directed acyclic graph (DAG) model, which abstracts construction procedures as nodes in the graph and represents the sequential relationship between procedures as directed edges. By analyzing node timestamps and procedure weights, mathematical formulas are used to determine whether the procedure chain conforms to the logical order. This technology can effectively detect problems such as reversed procedures and parallel procedure errors that may exist during construction, ensuring the standardization and rationality of the construction process. Its core lies in transforming complex logical relationships between procedures into computable mathematical expressions through quantification, thereby achieving automated verification.

[0095] IV. Module Workflow:

[0096] (a) Initialization phase:

[0097] After the four-dimensional dynamic coding engine is started, it loads the pre-set coding rules, process weight assignment rules, and relevant parameters of the topology constraint verification algorithm.

[0098] Establish communication connections with other modules of the system (such as the intelligent document verification module, dynamic traceability platform, etc.) to ensure timely acquisition of relevant information during the construction process and to transmit the generated codes and verification results to other modules;

[0099] (II) Encoding Generation Stage:

[0100] When a new construction procedure begins, the coding generation unit receives information such as the project type, structural parts, construction time, and spatial location from the construction site or other modules;

[0101] According to the coding rules, this information is combined to generate a unique code identifier, and the code is then bound to the process.

[0102] (III) Topology constraint verification phase:

[0103] The topology constraint verification unit obtains process chain information including process timestamps and process attributes;

[0104] The topological constraint verification algorithm is used to verify the process chain and determine the legality of the logical relationship between the processes.

[0105] The verification results are fed back to the system. If the verification passes, the process is allowed to continue; if the verification fails, an alert is issued and the relevant operation is suspended, pending adjustments by the construction personnel.

[0106] (iv) Encoding, storage, and management phase:

[0107] The encoding management and storage unit stores the generated encoding and related information in the database;

[0108] If any information changes during construction, the codes should be updated promptly, and the coding information in the database and relevant construction documents should be updated synchronously; at the same time, the code change history should be recorded.

[0109] (V) Conclusion:

[0110] When the entire construction project is completed or a stop command is received, the coding engine stops coding generation and verification operations, closes communication connections with other modules, backs up and archives relevant data during operation, and releases system resources.

[0111] V. Application Value of the Module:

[0112] (a) Achieving precise construction management:

[0113] Through unique coding and process logic verification, construction managers can clearly grasp the construction progress and process execution, promptly identify and correct missing documents during construction, ensure that construction is carried out in accordance with the predetermined plan and specifications, and improve the precision of construction management.

[0114] (ii) Improve document traceability efficiency:

[0115] The coding serves as an index for construction documents, closely linking them to the actual project. When it is necessary to consult construction documents or trace the construction process, the relevant documents and procedures can be quickly located through the coding, greatly improving the efficiency of document tracing and providing convenience for project acceptance, auditing, and other work.

[0116] (III) Ensuring construction safety and quality:

[0117] The topology constraint verification algorithm ensures the rationality of construction procedures and avoids safety hazards and quality problems caused by errors in procedures. At the same time, it assigns high weights to key procedures and safety control procedures for focused monitoring, further ensuring the safety and quality of the construction process.

[0118] (iv) Supporting the digital transformation of engineering:

[0119] The digital coding information generated by the four-dimensional dynamic coding engine provides basic data for the digital management of the construction process. Combined with technologies such as BIM models and dynamic traceability platforms, it can realize the visualization and intelligent management of the construction process and promote the digital transformation of the engineering construction industry.

[0120] In this embodiment, the intelligent file verification module executes the weighted Levenshtein distance algorithm:

[0121] Calculate the minimum difference value between the actual file set and each project template. This difference value consists of two parts:

[0122] The first part is the sum of the absolute value of the difference between the actual file status and the template file status multiplied by the file importance weight;

[0123] The second part is the product of the hierarchy decay factor and the hierarchy depth of the missing file in the project structure tree;

[0124] Among them, the importance weight of the document is assigned a value according to the criticality of the document, and the level depth corresponds to different values ​​for unit project, sub-item project and process level respectively;

[0125] Methods for building an engineering template library include:

[0126] Analyze the component characteristics of historical engineering drawings, decompose them according to the four-level structure from unit project to sub-item project to professional category, and generate a file tree corresponding to the coding level;

[0127] By linking construction specification tables with process sequence codes, a dynamically updatable template knowledge graph is formed;

[0128] Furthermore, the intelligent document verification module is one of the core components of the process-level multi-dimensional coding construction document inspection system. It acts like an "intelligent steward" for project document management, accurately identifying missing and abnormal documents based on an engineering template library and advanced algorithms. The following section elaborates on its functions, sub-modules, and technical principles:

[0129] I. Overall Function Overview:

[0130] The intelligent document verification module uses the engineering template library as a benchmark. By comparing and analyzing the actual construction documents with the template documents, it generates a document integrity matrix to determine whether there are any missing or incomplete construction documents. This module uses an improved Levenshtein distance algorithm to quickly match missing documents and output a list of replacement documents, ensuring the integrity and standardization of engineering documents. At the same time, the module can also dynamically adjust the verification strategy according to the progress of the project and the updates to the specifications, ensuring the accuracy and timeliness of document verification.

[0131] II. Submodule Composition and Functions:

[0132] (a) Process Document Template Library Management Unit:

[0133] Template Construction: By parsing the component characteristics of historical engineering drawings, a file tree is generated by breaking down the project into four levels: unit project → sub-project → itemized project → work process category. For example, in highway engineering, the overall project is first divided into unit projects such as bridge engineering and roadbed engineering. Then, bridge engineering is further subdivided into sub-projects such as foundation and substructure, and prefabrication and installation of superstructure. The foundation and substructure are further subdivided into itemized projects such as pile foundation and pile cap. Finally, the required construction documents, such as construction layout, concrete pouring records, and quality inspection reports, are compiled for each work process category.

[0134] Knowledge graph association: The construction specification table is associated with the process code to form a dynamically updated template knowledge graph; based on construction specifications such as GB50300, the document requirements corresponding to each process are determined and bound to the process code generated by the four-dimensional dynamic coding engine; when the construction specifications are updated, the template knowledge graph is automatically adjusted synchronously to ensure the timeliness and accuracy of the template library;

[0135] (II) Data Acquisition and Preprocessing Unit:

[0136] Data Acquisition: Real-time acquisition of various document information generated during construction, including but not limited to construction drawings, technical disclosure documents, material inspection reports, etc.; at the same time, acquisition of the coding information of the current construction process from the four-dimensional dynamic coding engine to clarify the project type, structural part and process sequence corresponding to the document;

[0137] Data preprocessing: Perform preprocessing operations such as format conversion and deduplication on the collected file data; convert files of different formats into a format that the system can recognize, remove duplicate file records, ensure the consistency and validity of input data, and provide a reliable data foundation for subsequent file verification;

[0138] (III) Intelligent File Verification Algorithm Unit:

[0139] Document Integrity Matrix Generation: Based on the project file template library, a corresponding document integrity matrix is ​​generated for each construction process; each element in the matrix corresponds to a file type, and the value indicates whether the file is required in the current process (1 indicates required, 0 indicates not required).

[0140] Weighted Levenshtein distance algorithm execution: using the formula

[0141] (in, This is the actual file state vector. This refers to the actual file state, when it exists. When missing ; For the template library file state matrix, For the template library The template's first Class file status; This is a file weight matrix, with key file weights. Weight of ordinary files ; The number of templates in the template library. The number of file types; This is the hierarchical attenuation factor, with a fixed value of 0.5; The algorithm calculates the difference between the actual file status and the template library file status, with a value of 4 for the unit project, 3 for the sub-project, 2 for the sub-item project, and 1 for the process level, representing the missing file level depth. This algorithm not only considers whether a file is missing, but also performs weighted calculations based on the importance and level of the file, highlighting the importance of key files and improving the accuracy of verification.

[0142] Missing document identification and list generation: Based on the calculation results of the weighted Levenshtein distance algorithm, the types of missing documents are determined and a replacement list is generated. The list includes information such as the name of the missing document, the work process to which it belongs, and its importance, so that construction personnel can promptly understand the missing document situation and replace it.

[0143] (iv) Verification Result Feedback and Management Unit:

[0144] Results Feedback: The document verification results will be fed back to relevant personnel in real time, including construction management personnel and document managers; reminders will be sent through various means such as system pop-ups, emails, and SMS to ensure that relevant personnel are aware of any missing documents in a timely manner;

[0145] Progress tracking: Track and manage the progress of missing document replacement, record information such as submission time and review status; when the document replacement is completed, automatically update the document integrity matrix and perform verification again to ensure that the document meets the requirements;

[0146] III. Key Technology Principles:

[0147] (I) Construction principle of the project document template library:

[0148] The engineering document template library is built based on knowledge engineering and information classification technology. Through the analysis of historical engineering data and the interpretation of construction specifications, engineering documents are systematically classified according to engineering structure and professional category. Using knowledge graph technology, construction specifications are associated with procedures and documents to form a knowledge network with semantic relationships. This structured template library can comprehensively and accurately reflect the document requirements under different engineering types and procedures, providing a reliable reference standard for document verification.

[0149] (II) Principle of Weighted Levenshtein Distance Algorithm:

[0150] The weighted Levenshtein distance algorithm is an improvement on the traditional Levenshtein distance algorithm, introducing file weights and a hierarchy decay factor. While the traditional Levenshtein distance algorithm calculates the difference between two strings, in file verification, the actual file state and the template file state are transformed into vector or matrix forms for calculation. Through the file weight matrix, critical files are given higher weights, making the impact of missing critical files on the distance calculation greater. The hierarchy decay factor considers the project hierarchy where the file resides; the lower the level of the file, the smaller the impact of missing files on the overall system. By reducing the weight of missing files through decay calculations, it better reflects the actual project situation and achieves accurate assessment of file integrity.

[0151] IV. Module Workflow:

[0152] (a) Initialization phase:

[0153] After the intelligent file verification module is started, it loads the latest version of the project template library, including the file tree structure and template knowledge graph;

[0154] Initialize the parameters of the weighted Levenshtein distance algorithm, such as the file weight matrix and the level decay factor, and establish communication connections with other modules (such as the four-dimensional dynamic encoding engine and the blockchain evidence storage and repair module).

[0155] (II) Data Collection and Preparation Stage:

[0156] The data acquisition and preprocessing unit collects construction document data and process code information in real time;

[0157] The collected data is preprocessed to ensure that the data format is consistent and the content is accurate;

[0158] (III) Document Verification Stage:

[0159] Based on the collected process code information, the corresponding file integrity matrix is ​​extracted from the project template library;

[0160] The weighted Levenshtein distance algorithm is used to compare the actual file status with the template file status to calculate the file integrity difference.

[0161] Based on the discrepancies, identify the missing documents and generate a replacement list;

[0162] (iv) Results Feedback and Follow-up Phase:

[0163] The verification result feedback and management unit will send the document verification results and the replacement list to the relevant responsible persons;

[0164] Continuously track the progress of replacing missing documents. Once the documents are replaced, re-verify them until all documents meet the requirements.

[0165] (V) Conclusion:

[0166] When the entire construction project is completed or a stop command is received, the module stops file verification operations, backs up and archives relevant data during operation, closes communication connections with other modules, and releases system resources.

[0167] V. Application Value of the Module:

[0168] (a) Ensure document integrity:

[0169] Through precise document verification and replacement list generation, missing construction documents can be identified and remedied in a timely manner, ensuring the integrity and standardization of project documents and providing strong support for project acceptance and auditing.

[0170] (ii) Improve management efficiency:

[0171] The automated document verification process replaces the traditional manual inspection method, which greatly reduces the time and labor costs of document inspection and improves the efficiency and accuracy of project document management.

[0172] (iii) Reducing compliance risks:

[0173] A library of engineering document templates closely linked to construction specifications ensures that document verification complies with relevant standards and requirements, avoids compliance risks caused by missing or non-standard documents, and guarantees the legal and compliant progress of engineering construction.

[0174] (iv) Supporting project collaboration:

[0175] Clear information on missing documents and a list of replacements facilitates information sharing and collaboration among all parties involved in the construction, promoting the smooth progress of the project.

[0176] In this embodiment, the blockchain evidence storage and repair module performs the following operations:

[0177] A lightweight Merkle tree verification algorithm is adopted, with a verification path complexity of logarithmic order.

[0178] The evidence storage data package includes the hash value of the previous block, a four-dimensional encoded identifier, the type of missing file, a timestamp, and the ID of the person responsible.

[0179] Data validation employs a two-factor authentication mechanism:

[0180] The first factor uses the Ed25519 algorithm to verify the signature of the combination of the encoded identifier and the timestamp;

[0181] The second factor uses the national cryptographic algorithm SM2 to verify the identity of the responsible person through signature verification;

[0182] Both verification results must pass for the data to be considered valid.

[0183] The time-limited completion mechanism includes a dynamic early warning strategy:

[0184] When a missing file is detected for the first time, a warning notification is pushed to the mobile terminal of the person in charge.

[0185] If not processed within 24 hours, the project chief engineer's terminal will be notified simultaneously.

[0186] If not handled within forty hours, an alarm at the project management system level will be triggered;

[0187] If the issue remains unresolved for more than 48 hours, an unalterable traceability certificate will be generated and the relevant acceptance process will be locked.

[0188] Furthermore, the blockchain-based evidence storage and repair module serves as the cornerstone of trust and the central hub for repair within the process-level multi-dimensional coding construction document inspection system. Through blockchain technology and intelligent algorithms, it achieves reliable evidence storage and efficient repair of missing construction document records. The following provides a comprehensive analysis from the aspects of functional positioning, module composition, technical principles, workflow, and application value:

[0189] I. Overall Function Overview:

[0190] The blockchain evidence storage and repair module is mainly responsible for storing missing records found during the construction document inspection process on the blockchain. It utilizes the immutable and traceable characteristics of blockchain to ensure the authenticity and credibility of the data. At the same time, the module has a built-in time-limited completion mechanism, which promotes the timely completion of missing documents through dynamic early warning strategies. If the missing documents are not repaired within the time limit, it triggers strict traceability and process locking to ensure the compliance and integrity of engineering document management.

[0191] II. Submodule Composition and Functions:

[0192] (a) Blockchain-based evidence storage unit:

[0193] Data collection for evidence storage: When the intelligent file verification module detects a missing file, the blockchain evidence storage unit immediately collects key information, including the four-dimensional code (engineering type code, structural part code, process sequence code, and spatiotemporal coordinate code) corresponding to the missing file, the type of the missing file, the evidence storage timestamp, and the responsible person ID (such as the construction team leader number).

[0194] Lightweight Merkle Tree Construction: A lightweight Merkle tree verification algorithm is used to process the evidence data, with a path complexity of O(log n). (in, To reduce the amount of evidence data, hash operations are used to compress a large amount of evidence data into a tree structure, ensuring fast data verification and efficient storage; for example, the verification time of 1,000 pieces of evidence data is reduced from linear complexity to logarithmic level, greatly improving verification efficiency.

[0195] Evidence storage data packet generation: Generate evidence storage data packets according to a fixed structure:

[0196] (in, This is the hash value of the previous block, used to link the blockchain; Encode the encrypted process steps to protect sensitive information; The file type is missing; For evidence storage timestamp; (for identifying the responsible party); the generated data packets are broadcast to blockchain nodes via a P2P network for consensus verification and storage;

[0197] Two-factor authentication is implemented: A two-factor authentication mechanism is used for the stored data, and the formula is as follows:

[0198] (in, Encoded-timestamped signatures generated based on the Ed25519 algorithm ensure data integrity and time validity. For the purpose of establishing a signature based on the national cryptographic SM2 algorithm to identify the responsible party and clarify the attribution of responsibility; This represents a logical AND operation; the stored data is only accepted if both signatures pass verification.

[0199] (II) Time-limited completion and early warning unit:

[0200] Countdown Start: Upon completion of evidence storage, a 48-hour completion countdown mechanism is initiated, using the evidence storage timestamp as the starting point to calculate the remaining completion time in real time;

[0201] Dynamic early warning strategy:

[0202] Initial missing documents (0-24 hours): The missing document information and resubmission requirements will be pushed to the responsible person via the APP, such as sending a notification that “[process code XXX + name] missing construction layout, please submit within 24 hours”;

[0203] 24-hour unresolved: The warning information will be synchronized to the project chief engineer's terminal, triggering a secondary reminder, and attaching a record of the responsible person's failure to handle the issue;

[0204] 40 hours without action: Triggers an alarm at the project management system level, sending alerts to relevant parties such as the supervision unit and the construction unit, indicating the risk of missing important documents;

[0205] If not repaired within 48 hours: Generate an unalterable traceability certificate, record the entire missing process, lock the acceptance process, and prohibit this process from entering the subsequent acceptance stage;

[0206] (III) Data Management and Traceability Unit:

[0207] Evidence storage data query: Supports quick query of blockchain evidence storage records by process code, responsible person ID, time range and other conditions, and provides a visual interface to display the content of evidence storage data package and blockchain verification path;

[0208] Traceability voucher generation: When a missing document issue enters the dispute or audit stage, traceability vouchers containing evidence data, early warning records, and processing logs are automatically generated, providing authoritative evidence for liability determination and dispute resolution;

[0209] Data synchronization and update: Real-time data synchronization with the dynamic traceability platform and intelligent document verification module ensures that each module has a consistent understanding of the missing document status, and updates the blockchain evidence status after the document is reissued, marking it as "repaired";

[0210] III. Key Technology Principles:

[0211] (I) Principles of Blockchain Evidence Preservation Technology:

[0212] Based on distributed ledger technology, missing records in a file are linked into a chain in chronological order as blocks. Each block contains the hash value of the previous block, forming an immutable chain structure. Merkle trees are used to efficiently verify data integrity, and data tampering can be quickly located by comparing hash values. Two-factor authentication combines the internationally used Ed25519 algorithm with the domestic cryptographic SM2 algorithm to achieve dual protection of data security and independent control, ensuring the legal validity of the stored data.

[0213] (II) Principle of Time-Limited Completion Mechanism:

[0214] The countdown and warning logic is implemented through timestamps and state machine models. The file missing status is divided into stages such as "initial missing - warning in progress - timeout without repair", with different processing strategies for each stage. Message queues and multi-terminal push technology are used to ensure that the warning information accurately reaches the relevant personnel. The process locking mechanism forces the problem to be resolved, forming a closed-loop management.

[0215] IV. Module Workflow:

[0216] (a) Initialization phase:

[0217] After the blockchain evidence storage and repair module is started, it loads the blockchain node configuration information, establishes a connection with the blockchain network, and synchronizes the latest block data.

[0218] Initialize the time-limited completion mechanism parameters, such as countdown duration, warning threshold, and list of responsible personnel, and complete the interface integration with other modules of the system (intelligent document verification module, dynamic traceability platform);

[0219] (ii) Evidence Triggering Stage:

[0220] When the intelligent file verification module detects a missing file, it sends the missing information to the blockchain evidence storage unit.

[0221] The evidence storage unit collects detailed data, constructs a Merkle tree and generates an evidence storage data package, which is then broadcast to the blockchain network for consensus storage after passing two-factor authentication.

[0222] (III) Time-limited completion and early warning stage:

[0223] A 48-hour countdown will be initiated, and a dynamic early warning strategy will be implemented according to the time nodes, pushing early warning information through multiple channels;

[0224] Real-time monitoring of the responsible person's handling status, recording the completion progress and operation logs;

[0225] (iv) Timeout processing stage:

[0226] If the file is not repaired within the timeout period, a traceability certificate is generated and the acceptance process is locked. At the same time, the final status is synchronized to all modules of the system.

[0227] (v) Repair Confirmation Phase:

[0228] After the responsible person submits the replacement documents, the intelligent document verification module re-verifies the integrity of the documents. Once the verification is successful, the blockchain evidence storage unit updates the evidence storage status to "repaired" and unlocks the acceptance process.

[0229] (vi) Conclusion:

[0230] After the project is completed, the module archives and stores the evidence data, closes the blockchain connection, and releases system resources.

[0231] V. Application Value of the Module:

[0232] (a) Ensuring data credibility:

[0233] Blockchain-based evidence storage ensures that missing document records are tamper-proof and fully traceable, providing authoritative evidence for project auditing and dispute resolution, and enhancing data credibility.

[0234] (II) Strengthen the implementation of responsibilities:

[0235] By clearly defining the responsible parties and implementing dynamic early warning systems, relevant personnel are compelled to address missing documents promptly, avoiding management loopholes caused by human error and improving project management execution.

[0236] (III) Preventing compliance risks:

[0237] The mechanism of locking the acceptance process after the deadline forces the project documents to meet the specifications, eliminates the phenomenon of "acceptance with defects", and reduces the compliance risks of project construction.

[0238] (iv) Optimize management processes:

[0239] Automated evidence storage, early warning, and traceability processes replace traditional manual ledger management, improve document management efficiency, and provide key support for the digital and intelligent transformation of engineering projects.

[0240] In this embodiment, the dynamic traceability platform includes:

[0241] Establish a two-way mapping relationship between two-dimensional construction drawings and BIM models, and use spatiotemporal coordinate codes as indexes to locate the component process progress;

[0242] An improved YOLOv5s model is used to identify coded regions in drawings, and its loss function consists of three parts.

[0243] The classification loss uses the focus loss function to address the class imbalance problem;

[0244] The target detection loss is multiplied by a coefficient of 0.8;

[0245] The encoding recognition enhancement loss is multiplied by a factor of 1.5, and this enhancement loss is specifically targeted at the four-dimensional encoded character region;

[0246] Bidirectional mapping is achieved through spatial coordinate transformation:

[0247] The spatiotemporal coordinate code is decomposed into two parts: the construction date code and the geographical coordinate abbreviation code;

[0248] The method for mapping geographic coordinate abbreviations to BIM model spatial grids is as follows:

[0249] Calculate the difference between the actual longitude and the longitude of the project origin, divide by the grid side length, and round down.

[0250] Calculate the difference between the actual latitude and the project origin latitude, divide by the grid side length, and round down.

[0251] Multiply the result calculated in the longitude direction by one thousand and add it to the result calculated in the latitude direction to obtain the unique identifier of the spatial grid;

[0252] Furthermore, the dynamic traceability platform, serving as the "digital monitoring hub" of the process-level multi-dimensional coded construction document inspection system, integrates BIM technology, computer vision, and intelligent algorithms to achieve visualized association and precise traceability of the entire lifecycle of construction documents and engineering entities. The following provides an in-depth analysis from the aspects of functional architecture, core modules, technical principles, workflow, and application value:

[0253] I. Overall Function Overview:

[0254] The dynamic traceability platform uses a unique code generated by a four-dimensional dynamic coding engine as a link to establish a two-way mapping relationship between construction drawings and BIM models, achieving precise association between construction documents, process progress, and engineering entities. By improving the YOLOv5s model recognition of drawing codes and combining spatiotemporal coordinate information, the platform enables real-time monitoring and dynamic traceability of the entire construction process. The platform supports the visual display of construction progress, document integrity status, and process logic relationships, providing intuitive and efficient decision support for engineering managers and ensuring transparency and traceability of the construction process.

[0255] II. Submodule Composition and Functions:

[0256] (a) BIM-coded bidirectional mapping unit:

[0257] Spatial coordinate transformation: The spatiotemporal coordinate code (6 bits) in the four-dimensional dynamic encoding is decomposed into a construction date code (3 bits) and a GPS coordinate abbreviation code (3 bits); among which, the GPS coordinate abbreviation code is generated using the formula... (in, , These are the actual latitude and longitude. , The coordinates of the project origin; The grid side length (default 10 meters) is mapped to the spatial grid of the BIM model (accuracy ±0.5m) to achieve precise alignment between the physical space and the digital model;

[0258] Two-way mapping construction: On the one hand, using codes as indexes, construction documents and process progress information are associated with corresponding components in the BIM model; on the other hand, by clicking on the components in the BIM model, the corresponding construction documents, process codes and historical progress data can be quickly retrieved, forming a two-way linkage query system of "model-code-file".

[0259] (ii) Encoded visual recognition unit:

[0260] Image acquisition and preprocessing: Using on-site deployed cameras, drones and other equipment, image data containing coded information such as construction drawings and component identification signs are acquired in real time; the acquired images are preprocessed by grayscale conversion, noise reduction and distortion correction to improve image quality and provide clear input data for code recognition;

[0261] Improved YOLOv5s Model Application: The improved YOLOv5s model is used for drawing code recognition, with the loss function being... (in, For classification loss, FocalLoss is used to handle the class imbalance problem; This is the target detection loss, used to determine whether a target exists in the encoded region; To enhance the recognition loss, the coded character region is assigned a weight of 1.5 times to improve the recognition accuracy of small-sized codes; the model quickly and accurately identifies the coded information in drawings by extracting and classifying the features of the coded characters.

[0262] Recognition result verification and output: The recognized code is verified by comparing it with the theoretical code generated by the four-dimensional dynamic coding engine to ensure the accuracy of the recognition; the verified code information is output to the platform database for subsequent construction progress tracking and document association.

[0263] (III) Dynamic monitoring and traceability unit:

[0264] Construction progress visualization: Using the BIM model as a carrier, the construction status of each component (such as not started, in progress, completed) is displayed through color, progress bars and other visualization methods based on the coded process progress information; managers can intuitively view the progress of the entire project from a three-dimensional perspective and quickly locate lagging processes.

[0265] Document traceability and early warning: Based on the results of the intelligent document verification module, the platform interface marks the process nodes where documents are missing in real time and issues early warning prompts; users can click on the early warning node to retrieve the list of missing documents and relevant responsible persons information with one click, so as to quickly locate and handle the problem;

[0266] Historical data backtracking: Supports backtracking the construction process by timeline, process code, and other conditions, allowing users to view the construction status, document version, and process logic at any point in time; through comparative analysis of historical data, it provides data support for project review and optimization;

[0267] (iv) Data Interaction and Integration Unit:

[0268] System Integration: Interacts with the four-dimensional dynamic encoding engine, intelligent file verification module, and blockchain evidence storage and repair module to obtain encoding information, file verification results, and evidence storage records in real time, ensuring the consistency and collaboration of data across modules;

[0269] Third-party system integration: Provides standard API interfaces to support integration with third-party systems such as project management systems and supervision platforms, enabling the sharing and collaboration of construction data, breaking down information silos, and improving project management efficiency;

[0270] III. Key Technology Principles:

[0271] (I) BIM-Encoding Two-Way Mapping Principle:

[0272] Based on spatial coordinate transformation algorithms and the parametric characteristics of BIM models, construction information in physical space is mapped to digital models; by establishing a unique correspondence between codes and model components, bidirectional indexing of data is achieved using database technology; when the actual construction progress or file status changes, the BIM model display status is synchronously driven to change by updating the code-related data, and vice versa.

[0273] (II) Principles of the Improved YOLOv5s Model:

[0274] Based on the YOLOv5s object detection framework, and considering the small size and dense arrangement of construction drawing codes, the model enhances its feature extraction capability for coded character regions by adjusting the weights of the loss function; FocalLoss reduces the weight of easily classified samples, making the model focus more on difficult-to-recognize codes; and for... By assigning higher weights, the model is guided to prioritize ensuring the accuracy of encoding recognition, thereby improving the overall recognition performance;

[0275] (III) Dynamic monitoring and traceability principle:

[0276] Real-time database technology is used to store dynamic data such as construction progress and file status in an in-memory database, enabling fast data reading, writing and updating; front-end visualization technologies such as WebGL and Three.js are used to render the data onto the BIM model, displaying the construction process in an intuitive interface; timestamp and version control technologies are used to achieve orderly storage and retrospective query of historical data.

[0277] IV. Module Workflow:

[0278] (a) Initialization phase:

[0279] After the dynamic traceability platform is started, it loads BIM model data and basic information such as project origin coordinates, and establishes data connections with other modules.

[0280] Initialize and improve the YOLOv5s model parameters, load pre-trained weights, and complete model deployment;

[0281] (II) Data Acquisition and Processing Stage:

[0282] The coded visual recognition unit collects construction image data and performs preprocessing;

[0283] The improved YOLOv5s model is used to identify the encoded information in the image, and the result is transmitted to the BIM-encoded bidirectional mapping unit after verification.

[0284] (III) Two-way mapping construction stage:

[0285] The BIM-coded bidirectional mapping unit maps the spatiotemporal coordinate information in the code to the spatial grid of the BIM model.

[0286] Establish the association between the coding and BIM model components and construction documents, and store them in the platform database;

[0287] (iv) Dynamic monitoring and traceability stage:

[0288] The dynamic monitoring and traceability unit acquires data from each module in real time, updates the visualization of the BIM model, and provides early warning information for missing annotation files.

[0289] Responding to user requests for data retrieval, querying and displaying historical data based on specific criteria;

[0290] (V) Data Interaction and Integration Phase:

[0291] The data interaction and integration unit receives data updates from other modules and synchronizes them to the platform database.

[0292] Data exchange with third-party systems is achieved through API interfaces to realize information sharing;

[0293] (vi) Conclusion:

[0294] After the project is completed, the platform archives and stores historical data, closes data connections, and releases system resources.

[0295] V. Application Value of the Module:

[0296] (a) Improve construction transparency:

[0297] By visually integrating BIM models with construction data, project managers can gain a comprehensive understanding of construction progress and document status, reducing information asymmetry and improving decision-making efficiency.

[0298] (II) Strengthen quality traceability capabilities:

[0299] The coding-based full lifecycle traceability function can quickly locate the source of construction problems, facilitate the analysis and rectification of quality problems, and improve the level of project quality control.

[0300] (III) Promoting multi-party collaborative management:

[0301] Integration with third-party systems enables the sharing of construction data, promotes collaborative work among construction units, construction companies, supervision units and other parties, and improves the coordination and efficiency of project management.

[0302] (iv) Promote the digital transformation of engineering:

[0303] As a core tool for digital engineering management, the dynamic traceability platform upgrades the traditional construction management model towards intelligence and visualization, providing technical support for the digital transformation of the construction industry.

[0304] A method for detecting multi-dimensional coded construction documents at the process level. This method utilizes a process-level multi-dimensional coded construction document detection system to detect process-level multi-dimensional coded construction documents. The method includes the following steps:

[0305] Encoding Generation and Logic Verification: Utilizing a four-dimensional dynamic encoding engine, a unique encoding identifier is generated for each construction process based on information such as project type, structural location, process sequence, and spatiotemporal coordinates. Simultaneously, a topological constraint verification algorithm, based on an improved directed acyclic graph (DAG) model, verifies the legality of the process logic relationships. The difference in timestamps between process nodes in the process chain is calculated, and combined with process weight factors (basic process weight is 1, key quality control point process weight is 2, and safety control process weight is 2.5), it is determined whether it meets the legality threshold (valued as the process chain length). If it does, the process logic is legal; otherwise, an alert is issued and the relevant file flow is blocked.

[0306] Document integrity verification: The intelligent document verification module is based on the engineering template library. It generates a document tree by breaking down the document into four levels: unit project → sub-project → sub-item project → work process category, and associates it with construction specifications and work process codes to form a knowledge graph. It collects document data during the construction process, performs preprocessing, and uses a weighted Levenshtein distance algorithm to compare the actual document status with the document status in the template library. It considers the document weight matrix (key documents have a weight of ≥2, and ordinary documents have a weight of 1) and the hierarchical decay factor (fixed value of 0.5) to calculate the difference between the two, identify missing documents, and generate a replacement list.

[0307] Missing Document Preservation and Repair: When a missing document is detected, the blockchain preservation and repair module is activated. It collects information related to the missing document, including its encoding, missing file type, timestamp, and responsible person ID. A lightweight Merkle tree verification algorithm is used to construct the preservation data package, and two-factor authentication (based on Ed25519 encoding-timestamp signature and SM2-based responsible person identity signature) ensures the credibility of the preserved data. Simultaneously, a 48-hour completion countdown is initiated, with dynamic warnings issued according to time nodes (initial missing document is pushed to the responsible person via the APP; if not processed within 24 hours, it is synchronized to the project chief engineer's terminal; if not processed within 40 hours, a system-level alarm is triggered). If the document is not repaired within the time limit, a traceability certificate is generated and the acceptance process is locked.

[0308] Dynamic traceability throughout the entire construction process: The dynamic traceability platform constructs a two-way mapping between construction drawings and BIM models, decomposing spatiotemporal coordinate codes into construction date codes and GPS coordinate abbreviations, and mapping them to the BIM model spatial grid through a specific formula (accuracy ±0.5m); it utilizes an improved YOLOv5s model to recognize drawing codes, and its loss function enhances the recognition accuracy of coded character areas; it monitors construction progress in real time, visually displays the construction status of each component on the BIM model, and annotates document missing warning information; it supports backtracking construction history data according to time axis, process code, and other conditions, realizing full lifecycle visual traceability of construction documents and engineering entities;

[0309] Data interaction and closed-loop management: Real-time data interaction between modules ensures consistency of data such as coding information, file verification results, and evidence storage records. For example, the intelligent file verification module transmits missing file information to the blockchain evidence storage and repair module, and the dynamic traceability platform obtains the coding and file status for visualization. The entire detection process forms a closed-loop management system, from file detection, problem evidence storage, repair tracking to result traceability, ensuring the standardization and integrity of construction document management.

[0310] While 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 these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A process-level multi-dimensional coding construction document inspection system, characterized in that: include: The four-dimensional dynamic coding engine generates a unique four-dimensional coding identifier consisting of an engineering type code, a structural part code, a process sequence code, and a spatiotemporal coordinate code. It then verifies the legality of the process logic relationships using a topological constraint verification algorithm based on an improved directed acyclic graph model. The verification process is as follows: Calculate the sum of the products of the time difference function values ​​of all adjacent process nodes in the process chain to be verified and the process weight factor. When the sum is greater than or equal to the validity threshold, output a verification pass signal; otherwise, output a verification failure signal. The time difference function outputs a value of 1 when the timestamp of the subsequent process is greater than the timestamp of the preceding process, and outputs a value of 0 otherwise. Process weighting factors are assigned values ​​according to the importance of each process. The validity threshold is equal to the total number of nodes in the process chain; Intelligent file verification module: Based on a pre-set project file template library, it generates a file integrity matrix, uses an improved Levenshtein distance algorithm to match missing files and outputs a replacement list. The intelligent file verification module executes a weighted Levenshtein distance algorithm: Calculate the minimum difference value between the actual file set and each project file template. This difference value consists of two parts: The first part is the sum of the absolute value of the difference between the actual file status and the template file status multiplied by the file importance weight; The second part is the product of the hierarchy decay factor and the hierarchy depth of the missing file in the project structure tree; Among them, the importance weight of the document is assigned a graded value according to the criticality of the document, and the hierarchical depth corresponds to different values ​​for unit project, sub-project, sub-item project and process level respectively. Blockchain Evidence Preservation and Repair Module: Preserves missing file records on the blockchain and triggers a timed completion mechanism; Dynamic Traceability Platform: Connects a four-dimensional dynamic coding engine with an intelligent document verification module to achieve visualized monitoring of the entire lifecycle of construction documents.

2. The process-level multi-dimensional coding construction document detection system according to claim 1, characterized in that: The method for constructing the project file template library includes: Analyze the component characteristics of historical engineering drawings, decompose them according to the four-level structure of unit project to sub-project to sub-item project and then to process category, and generate a process file tree corresponding to the coding level; By linking the construction specification table with the process sequence code, a dynamically updatable template knowledge graph is formed.

3. The process-level multi-dimensional coding construction document detection system according to claim 1, characterized in that: The blockchain evidence storage and repair module performs the following operations: A lightweight Merkle tree verification algorithm is adopted, with a verification path complexity of logarithmic order. The evidence storage data package includes the hash value of the previous block, a four-dimensional encoded identifier, the type of missing file, a timestamp, and the ID of the person responsible. Data validation employs a two-factor authentication mechanism: The first factor uses the Ed25519 algorithm to verify the signature of the combination of the encoded identifier and the timestamp; The second factor uses the national cryptographic algorithm SM2 to verify the identity of the responsible person through signature verification; Both verification results must pass for the data to be considered valid.

4. The process-level multi-dimensional coding construction document detection system according to claim 1, characterized in that: The dynamic traceability platform includes: Establish a two-way mapping relationship between two-dimensional construction drawings and BIM models, and use spatiotemporal coordinate codes as indexes to locate the component process progress; An improved YOLOv5s model is used to identify coded regions in drawings, and its loss function consists of three parts. The classification loss uses the focus loss function to address the class imbalance problem; The target detection loss is multiplied by a coefficient of 0.8; The encoding recognition enhancement loss is multiplied by a factor of 1.5, and this enhancement loss is applied to the four-dimensional encoded character region.

5. The process-level multidimensional coding construction document detection system according to claim 1, characterized in that: The assignment rules for the process weighting factors are as follows: Basic process types include template installation and measurement and layout, which are assigned benchmark weight values; Key quality control points include concrete pouring and prestressing tensioning, which are given a weight twice that of the baseline value. Safety control procedures include foundation pit support and blasting operations, which are assigned a weight 2.5 times that of the baseline value.

6. The process-level multidimensional coding construction document detection system according to claim 1, characterized in that: The time-limited completion mechanism includes a dynamic early warning strategy: When a missing file is detected for the first time, a warning notification is pushed to the mobile terminal of the person in charge. If not processed within 24 hours, the project chief engineer's terminal will be notified simultaneously. If not handled within forty hours, an alarm at the project management system level will be triggered; If the issue remains unresolved for more than 48 hours, an unalterable traceability certificate will be generated and the relevant acceptance process will be locked.

7. The process-level multi-dimensional coding construction document detection system according to claim 4, characterized in that: The bidirectional mapping is achieved through spatial coordinate transformation: The spatiotemporal coordinate code is decomposed into two parts: the construction date code and the geographical coordinate abbreviation code; The method for mapping geographic coordinate abbreviations to BIM model spatial grids is as follows: Calculate the difference between the actual longitude and the longitude of the project origin, divide by the grid side length, and round down. Calculate the difference between the actual latitude and the project origin latitude, divide by the grid side length, and round down. Multiply the result calculated in the longitude direction by 1000 and add it to the result calculated in the latitude direction to obtain the unique identifier of the spatial grid.

8. A method for detecting construction documents with multi-dimensional coding at the process level, characterized in that: The method utilizes the process-level multidimensional coded construction document detection system described in any one of claims 1-7 to detect process-level multidimensional coded construction documents.

Images