A BIM-based engineering information management system

Abstract

The application discloses a BIM-based engineering information management system, and relates to the field of building engineering information management.The system comprises a data processing module, a model construction module, a blockchain storage module, a verification module, a data fusion module and a permission management module.The data processing module receives original engineering data and generates a structured data signal.The model construction module generates a BIM model according to the structured data signal.The blockchain storage module encrypts the metadata signal and stores the same, and generates a storage completion signal which is fed back to the model construction module.The verification module analyzes a dynamic operation log signal, and generates a risk early warning signal and a repair suggestion signal when an abnormality is detected.The data fusion module generates a fusion data signal which is transmitted to the model construction module and triggers BIM model updating.The permission management module adjusts user permissions according to the risk early warning signal.The application can solve the problems of data silos, security risks, low collaboration efficiency and insufficient intelligence in engineering information management.

Description

Technical Field

[0001] This invention relates to the field of information management of building engineering, and specifically to a BIM-based engineering information management system. Background Technology

[0002] Building Information Modeling (BIM) technology has become a core tool in modern engineering management, significantly improving project efficiency through features such as 3D modeling, multidisciplinary collaboration, and construction simulation. However, with the expansion of project scale and the acceleration of digital transformation, the shortcomings of traditional BIM systems in data management, security, and intelligence are becoming increasingly apparent. First, project data is scattered across different professional teams and software platforms, and asynchronous model version updates frequently lead to data silos. For example, when structural design and MEP installation teams modify the same model node, version conflicts arise due to the lack of real-time collaboration mechanisms, requiring a significant amount of time to manually merge the data, severely hindering construction progress.

[0003] Secondly, BIM models typically employ a centralized storage architecture, which carries the risk of single points of failure. Furthermore, their access control systems are often rudimentary, making critical data susceptible to tampering or leakage. For instance, unauthorized modifications to model parameters by the construction company could lead to potential quality issues, but traditional auditing methods lack sufficient traceability to quickly pinpoint the responsible party. Additionally, model compliance checks heavily rely on human experience; problems such as pipeline collisions and structural overruns are often only discovered during construction, resulting in extremely high rework costs. Moreover, emerging real-time data from IoT sensors and drone point clouds lack deep integration with BIM models, hindering dynamic optimization of construction plans. Existing improvement solutions include some patents attempting to introduce blockchain technology to store critical model data, but this only addresses static metadata storage and does not cover the integrity verification of dynamic operation logs. Other solutions employ machine learning for model validation, but the AI ​​algorithms are disconnected from business processes, failing to link access control or trigger automatic repairs. Furthermore, multi-source data fusion often relies on customized interfaces, lacking a universally compatible architecture, which limits system scalability and makes it difficult to be compatible with sensors or GIS platforms from different vendors.

[0004] In terms of access control, traditional role-based classification cannot adapt to the dynamic needs of complex engineering scenarios, such as the timeliness control of temporary contractor access permissions or the fine-grained data sharing needs in cross-regional team collaboration. These combined issues make it difficult for existing BIM systems to support refined management throughout the entire project lifecycle. There is an urgent need for an innovative solution that integrates trusted data storage, intelligent verification, and dynamic collaboration to achieve efficient cross-platform, multi-role collaboration and real-time decision-making while ensuring data security. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a BIM-based engineering information management system to solve the problems of data silos, security risks, low collaboration efficiency, and insufficient intelligence in engineering information management. This invention ensures data trustworthiness through blockchain encryption and notarization, verifies model defects in real time using AI, improves collaboration efficiency through dynamic fusion of multi-source data, and achieves fine-grained access control through federated learning.

[0006] This invention provides a BIM-based engineering information management system, comprising:

[0007] The data processing module receives raw engineering data and generates structured data signals.

[0008] The model building module generates a BIM model based on structured data signals and extracts model metadata signals and dynamic operation log signals.

[0009] The blockchain evidence storage module encrypts and stores metadata signals and generates a evidence storage completion signal that is fed back to the model building module.

[0010] The verification module analyzes dynamic operation log signals and generates risk warning signals and repair suggestion signals when an anomaly is detected.

[0011] The data fusion module integrates model metadata signals, dynamic operation log signals, and repair suggestion signals to generate a fused data signal, which is then transmitted to the model building module and triggers a BIM model update.

[0012] The permission management module adjusts user permissions based on risk warning signals, generates operation interception signals and sends them to the federated learning nodes to train the abnormal access model. After training, the federated learning nodes send permission policy update signals to the permission management module and synchronize them to the blockchain evidence storage module for record keeping.

[0013] In one embodiment of the present invention, the data processing module includes a data cleaning unit and a format conversion unit. The data cleaning unit filters redundant fields in the original engineering data using a preset regular expression template and a machine learning model, and generates a cleaned intermediate data signal. After receiving the intermediate data signal, the format conversion unit converts the unstructured data into a uniformly encoded structured data signal according to the BIM model construction standard. The structured data signal includes geometric parameters, material properties, and construction progress tags. After completing the conversion, the format conversion unit sends a data ready signal to the model construction module and simultaneously encrypts and transmits the metadata of the structured data signal to the blockchain notarization module for pre-notarization.

[0014] In one embodiment of the present invention, the model building module is equipped with a parametric modeling engine. The parametric modeling engine generates a three-dimensional model skeleton based on the geometric parameters in the structured data signal, and loads the corresponding physical property calculation model through material properties to generate the geometric property signal and material parameter signal of the BIM model. At the same time, it records the user's editing operations on the model to form a dynamic operation log signal. The dynamic operation log signal includes the operation timestamp, user ID and hash value of the modified content. When concurrent editing operations on the same model node are detected, the model building module sends a version conflict warning signal to the blockchain evidence storage module and triggers a version lock signal to suspend the editing permission of the current node.

[0015] In one embodiment of the present invention, the blockchain evidence storage module uses an asymmetric encryption algorithm to fragment the metadata signal, stores the fragmented data blocks to a private chain node and a distributed file system respectively, and writes the hash value of each data block into the blockchain ledger to form an evidence storage chain. The version lock signal is implemented through a smart contract. When more than three related metadata blocks are written consecutively into the evidence storage chain, the smart contract automatically generates a version lock signal and transmits it to the model building module. The version lock signal is only revoked by the blockchain evidence storage module after receiving the risk release confirmation signal sent by the verification module.

[0016] In one embodiment of the present invention, the verification module has a built-in defect identification model and a rule engine. The defect identification model is trained on a historical engineering defect dataset and is used to identify pipeline collisions, structural overruns, or specification conflicts in dynamic operation log signals. The rule engine loads an industry standard specification library and compares it with the BIM model attributes in real time. When an anomaly is detected, the verification module generates a risk warning signal containing risk level, location coordinates, and repair priority labels. At the same time, it calls a pre-set simulation engine to simulate a repair plan and generates a repair suggestion signal containing parameter adjustment suggestions, alternative path planning, and cost impact analysis. The repair suggestion signal triggers the data fusion module to start the model update process.

[0017] In one embodiment of the present invention, the data fusion module includes a spatiotemporal alignment unit and a conflict resolution unit. The spatiotemporal alignment unit normalizes the timestamps of IoT (Internet of Things) sensor data, UAV point cloud data, and GIS (Geographic Information System) data, and maps them to the corresponding positions in the BIM model through spatial coordinate system transformation to generate a spatiotemporally synchronized fused data signal. When the conflict resolution unit detects that the deviation between the sensor data and the preset threshold of the BIM model exceeds 10%, it automatically generates a data anomaly report and triggers the secondary verification of the verification module. The fused data signal drives the model building module to iteratively optimize the material parameters or structural topology of the BIM model, and at the same time sends a real-time update command to the visualization module.

[0018] In one embodiment of the present invention, the access control module includes a geofencing engine and a behavior analysis model. The geofencing engine generates a dynamic electronic fence based on the construction area boundary in the GIS data. When the GPS (Global Positioning System) coordinates of the user terminal exceed the fence range, the user's data access permissions are automatically downgraded and an operation interception signal is generated. The behavior analysis model obtains encrypted user operation history data through federated learning nodes and trains an abnormal access detection model. When an abnormal operation sequence is detected, the access control module dynamically restricts the user's editing permissions for key nodes of the BIM model and transmits the interception log to the blockchain evidence storage module in encrypted form.

[0019] In one embodiment of the present invention, the federated learning node uses homomorphic encryption technology to aggregate user behavior data from multiple terminals. After training an abnormal access detection sub-model locally, the model gradient parameters are uploaded to the central server for aggregation and update, generating a global permission policy update signal. The permission policy update signal includes a user role permission matrix, a data sensitivity classification table, and an operation blacklist and whitelist. After receiving the update signal, the permission management module initiates secondary authentication for users with security risks exceeding the threshold and writes the updated policy version hash value into the audit-specific chain of the blockchain evidence storage module.

[0020] In one embodiment of the present invention, the model building module further includes a dynamic version merging unit. After receiving the version locking signal sent by the blockchain evidence storage module, the dynamic version merging unit automatically parses the concurrent modification records of multiple users on the same BIM model node, compares the differences in geometric properties, material parameters and related logic in each version through a conflict resolution algorithm, and generates a version integration scheme containing weight scores and merging priority labels.

[0021] In one embodiment of the present invention, after the integration scheme verifies the structural stability through a lightweight simulation engine, it triggers the model building module to generate an updated BIM model version, and transmits the merge log in encrypted form to the blockchain evidence storage module to generate a new evidence storage chain branch. At the same time, it sends a version merge completion signal to the permission management module to remove the editing operation restriction.

[0022] This invention provides a BIM-based engineering information management system that ensures data credibility through blockchain encryption and notarization, verifies model defects in real time with AI, improves collaboration efficiency through dynamic fusion of multi-source data, and achieves fine-grained access control through federated learning. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a system architecture diagram of a BIM-based engineering information management system. Detailed Implementation

[0025] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0026] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0027] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0028] Please see Figure 1 The diagram illustrates a BIM-based engineering information management system according to the present invention. This system includes a data processing module, a model building module, a blockchain notarization module, a verification module, a data fusion module, and a permission management module. The data processing module receives raw engineering data and generates structured data signals. The model building module generates a BIM model based on the structured data signals and extracts model metadata signals and dynamic operation log signals. The blockchain notarization module encrypts and stores the metadata signals and generates a notarization completion signal, which is fed back to the model building module. The verification module analyzes the dynamic operation log signals and generates risk warning signals and repair suggestion signals when anomalies are detected. The data fusion module integrates the model metadata signals, dynamic operation log signals, and repair suggestion signals to generate a fused data signal, which is transmitted to the model building module and triggers a BIM model update. The permission management module adjusts user permissions based on the risk warning signals, generates operation interception signals, sends them to the federated learning nodes to train an abnormal access model, and after training, the federated learning nodes send permission policy update signals to the permission management module and synchronize them to the blockchain notarization module for record-keeping.

[0029] like Figure 1 As shown, the data processing module serves as the system entry point, receiving raw engineering data from user terminals (such as design software and mobile devices) or external devices (such as sensors and drones), including design drawings, material lists, construction logs, and point cloud scan data. This module preprocesses the raw data through a data cleaning unit, specifically by using regular expression templates to match redundant information in unstructured text (such as duplicate records or invalid symbols in construction logs) and using machine learning models (such as LSTM-based sequence classifiers) to identify outliers in the data (such as concrete strength values ​​exceeding reasonable ranges), generating cleaned intermediate data signals. Subsequently, the format conversion unit converts the intermediate data into structured data signals conforming to BIM standards. For example, it maps layer information in CAD drawings to geometric parameters in IFC (Industrial Basic Class) format and converts Excel material tables into standardized attribute tables containing strength grades and supplier codes. The structured data signals not only contain static information such as geometric parameters and material properties but also include construction stage labels (such as "foundation construction period" and "steel structure installation period") for subsequent model version management. After the conversion is completed, the data processing module sends a data ready signal to the model building module and simultaneously encrypts and transmits the metadata of the structured data (such as data source and conversion timestamp) to the blockchain notarization module for pre-notarization to ensure the credibility of data traceability.

[0030] Furthermore, the model building module generates a BIM 3D model based on structured data signals, with its core being a parametric modeling engine. This engine first parses geometric parameters (such as beam and column cross-sectional dimensions, and pipe routing coordinates) to construct a 3D model skeleton. Then, it loads material parameters (such as the elastic modulus of steel and the compressive strength of concrete) and associates them with physical property calculation models (such as finite element analysis models), generating a complete BIM model containing geometric attribute signals (vertex coordinates, topological relationships) and material parameter signals (mechanical properties, durability indicators). During model editing, the module records user operations (such as component movement and attribute modification) in real time, forming dynamic operation log signals. Each log entry includes an operation timestamp, user ID, a hash value of the modified content (e.g., generated using the SHA-256 algorithm), and an impact scope marker (e.g., "Affects only the electrical system"). When multiple users concurrently edit the same model node (such as fire hydrants on a certain floor), the model building module sends a version conflict warning signal to the blockchain notarization module and triggers a version lock signal to suspend the editing permissions of that node until the conflict is resolved. In addition, the module incorporates a lightweight rendering engine to convert BIM models into a lightweight format (such as GLTF) that can be displayed on mobile devices, while retaining key attribute information for subsequent verification analysis. The blockchain evidence storage module is responsible for ensuring the immutability and traceability of BIM data. This module adopts a hybrid storage architecture: for model metadata signals (such as geometric attributes and material parameters), asymmetric encryption algorithms (such as RSA-2048) are used for sharding, storing data blocks separately on private blockchain nodes (deployed on internal servers of the engineering company) and distributed file systems (such as IPFS), and writing the hash value of each data block to the blockchain ledger to form an evidence storage chain; dynamic operation log signals are processed by batch hashing using a Merkle tree structure, with the root hash periodically (e.g., every 10 minutes) uploaded to the blockchain. When three or more related metadata blocks (such as design change records for the same construction phase) are continuously written into the evidence storage chain, the module's built-in smart contract automatically generates a version lock signal, which is transmitted to the model building module to freeze the editing permissions of the current model version. Version locks are only released in two situations: first, when the verification module sends a risk release confirmation signal (e.g., defect fixes are completed); and second, when a manually authorized party (e.g., the project's chief engineer) unlocks the lock through a multi-signature verification mechanism. Furthermore, this module provides an audit interface, supporting the retrieval of historical evidence records by time range, user ID, or operation type, and generating legally valid audit reports.

[0031] In one embodiment of the present invention, the technical details of the data processing module are further defined, wherein the data cleaning unit and the format conversion unit constitute a two-stage preprocessing pipeline. The data cleaning unit first matches noise information in the original data using regular expression templates, such as filtering out non-standard date formats (e.g., "2023 / 13 / 32") or invalid characters (e.g., garbled symbols) in construction logs. Then, it uses a pre-trained LSTM model to identify abnormal data: the model input is a text sequence (e.g., a material testing report), and the output is an abnormal probability value. When the probability exceeds a threshold (e.g., 0.9), the data is automatically marked and notified for manual review. The format conversion unit adopts a combination of rule engine and machine learning: for highly structured data (e.g., CAD drawings), it directly converts the data using predefined IFC mapping rules (e.g., mapping "Layer1" to "structural beam"); ​​for unstructured data (e.g., handwritten notes by construction workers), it uses OCR recognition and inputs the data into a BERT model to extract key entities (e.g., "concrete pouring thickness 30cm"), and then converts them into standard attribute fields. During the conversion process, the module generates a data lineage map in real time, recording the mapping relationship between the original data and the structured data. This map is encapsulated in JSON-LD format and encrypted before being transmitted to the blockchain evidence storage module, providing a basis for subsequent data traceability. To address version conflict issues in the model building module, a dynamic version merging mechanism is proposed. When the blockchain evidence storage module triggers a version lock signal, the dynamic version merging unit within the model building module is activated. It first extracts concurrent editing records from the blockchain ledger, analyzing the differences in geometric attributes, material parameters, and related logic among the versions (e.g., user A modifies the pipe diameter to 200mm, and user B adds a support at the same location). The conflict resolution algorithm generates weighted scores based on preset priority rules (e.g., the designer's authority is higher than the construction party's) and physical feasibility assessments (e.g., finite element analysis verifies structural strength), and labels the merge priority (e.g., user A's compliance score is 85, user B's is 70), and marks the merge priority with tags such as "must retain" and "optional optimization." After the integrated solution is verified by a lightweight simulation engine (e.g., fluid simulation to verify flow loss after pipeline modification), the model building module is triggered to generate a new version of the BIM model. Simultaneously, the merge log (including conflict nodes, resolution logic, and final parameters) is encrypted and transmitted to the blockchain evidence storage module, generating an evidence storage chain branch (e.g., "V2.1_merge") for subsequent auditing and traceability. Furthermore, after the version merge is complete, the module sends a signal to the permission management module to remove editing restrictions on relevant users and updates their operation permission weights based on the merge results (e.g., downgrading the permissions of users who frequently cause conflicts).

[0032] like Figure 1As shown, the technical implementation of the verification module is further defined. This module integrates a dual verification mechanism of a deep learning-based defect identification model and an industry standard-driven rule engine. The defect identification model is trained using a historical engineering defect dataset, which covers typical problems such as pipeline collisions, excessive structural stress, and violations of fire separation distances. The model adopts a hybrid architecture of convolutional neural networks and graph neural networks, which can analyze the geometric topological relationships and physical property associations of the BIM model. For example, it can detect the collision risk of pipeline intersections at different elevations in the pipeline system, or assess the deformation threshold of steel structure nodes under load conditions. The rule engine has a built-in scalable industry standard library, including building fire protection codes, seismic design standards, and green building evaluation indicators. The engine compares the BIM model parameters with the code requirements in real time. When a deviation is detected, a risk warning signal containing risk level, location coordinates, and repair priority labels is generated. For example, if the width of the escape route is insufficient, it is marked as red high risk and associated with the evacuation simulation results. After identifying anomalies, the verification module invokes a lightweight simulation engine to simulate repair solutions, such as automatically adjusting pipeline elevations or adding structural supports. This generates a repair suggestion signal that includes parameter adjustment ranges, alternative path planning, and cost impact analysis. This signal triggers the data fusion module to initiate a model update process and simultaneously sends a risk level change notification to the access control module to adjust the corresponding user's operating permissions. The verification module also supports manual review. When a repair suggestion is rejected by a user, the module records the reason for the decision and reassesses the risk coefficient. If manual intervention escalates the risk, a secondary verification is forcibly triggered, and the editing permissions of the relevant model are frozen until the risk is resolved. The spatiotemporal alignment and conflict resolution capabilities of the data fusion module are further refined. This module includes a spatiotemporal alignment unit and a multi-source data conflict resolution unit. The spatiotemporal alignment unit first performs timestamp normalization on real-time data from IoT sensors, point cloud data collected by drones, and GIS geographic information data. It then synchronizes the clocks of all devices using a network time protocol and unifies the time base of all data to the main timeline of the BIM model. In the spatial coordinate alignment stage, a coordinate system transformation algorithm is used to convert the spatial reference system of data from different sources to the engineering coordinate system used in the BIM model. For example, GPS WGS-84 coordinates are converted to the construction local coordinate system. Point cloud data is then matched with the geometric surface of the BIM model using an ICP registration algorithm. The conflict resolution unit monitors deviations between the fused data and preset thresholds in the BIM model. For example, when the concrete curing temperature monitored by sensors exceeds the preset range of the model, the unit automatically triggers a data reliability assessment process. It corrects the original data by combining sensor calibration records and environmental factors. If significant deviations still exist after correction, a data anomaly report is generated and submitted to the verification module for secondary verification. Simultaneously, the automatic updating of relevant model parameters is paused.The data fusion module also drives the model building module to iteratively optimize the BIM model. For example, it can adjust the stiffness parameters of the foundation model based on real-time settlement monitoring data, or update the construction progress simulation logic based on meteorological data. The optimized model parameters generate a new version of the evidence record through the blockchain evidence storage module, and the difference is compared with the original version for audit traceability.

[0033] like Figure 1 As shown, the detailed specifications outline the collaborative control mechanism between geofencing and behavior analysis in the access control module. The geofencing engine generates a dynamic electronic fence based on the boundary coordinates of the construction area in GIS data. The fence's range automatically adjusts as construction progresses; for example, it's limited to 50 meters around the foundation pit during the foundation construction phase and expands to the entire building outline during the main structure phase. When a user's GPS coordinates exceed the fence's range, the engine downgrades their data access permissions in real time, converting their original design editing permissions to read-only mode and generating an operation interception signal to interrupt any sensitive operations being performed, such as model parameter modifications or drawing export requests. The behavior analysis model obtains encrypted user operation history data through federated learning nodes and uses a time-series pattern mining algorithm to identify abnormal access sequences, such as frequent switching of access to model nodes within a short period or attempts to download data from unauthorized areas. When abnormal behavior is detected, the model dynamically adjusts the user's permission level, such as restricting their editing permissions for key structural nodes or extending the operation approval process. The access control module also integrates a multi-factor authentication interface. When a user triggers a high-risk operation (such as deleting a historical version or exporting a complete model), biometric verification or dynamic token confirmation is required. All intercepted operations and access control change records are encrypted and transmitted to the blockchain evidence storage module to generate an immutable audit log. The log includes the operation time, device fingerprint, and associated risk score.

[0034] Furthermore, the workflow and encryption strategy of the federated learning nodes are defined. The federated learning nodes employ homomorphic encryption to process user behavior data uploaded from each terminal, ensuring that the data participates in model training under encrypted conditions. Each terminal trains an abnormal access detection sub-model locally, adds noise to the gradient parameters using differential privacy technology, and then uploads it to the central server. The server aggregates the gradients from each terminal to update the global model parameters, generating a global permission policy update signal that includes a user role permission matrix, a data sensitivity classification table, and operation blacklists and whitelists. After the permission policy update signal is sent to the permission management module, the module adjusts user permissions according to the latest policy. For example, it reduces the data download frequency limit for users who frequently trigger low-risk alarms, or enables real-time screen recording monitoring for high-risk roles. The node also sets model update trigger conditions. When a new attack pattern is detected (such as abnormal access from a specific IP segment) or a change in the project phase (such as transitioning from the design phase to the construction phase), an emergency training process is automatically initiated to generate a temporary permission policy. The validity period of the temporary policy is bound to the associated project milestone, and it automatically rolls back to the baseline permission settings after the policy expires. The federated learning nodes are deeply integrated with the blockchain evidence storage module's audit chain. Each policy update generates a digital fingerprint and writes it to the blockchain, ensuring that the history of permission changes is traceable and non-repudiable. It also supports third-party auditing institutions verifying policy compliance through authorized interfaces. Through refined technical design, the system's defect identification accuracy, multi-source data fusion reliability, dynamic permission control intelligence, and federated learning security are enhanced, forming a closed-loop management system from data verification to access control. The verification module's dual verification mechanism reduces the need for manual intervention; the data fusion module's spatiotemporal alignment and conflict resolution ensure the consistency of multi-source data; the permission management module's geofencing and behavioral analysis enable adaptive scenario control; and the federated learning nodes improve the generalization ability of the anomaly detection model while protecting privacy. The modules interact through signals to form a linkage mechanism. For example, an anomaly report from the data fusion module triggers a secondary verification in the verification module. The verification result further drives permission policy adjustments, and finally, blockchain evidence storage ensures the integrity and auditability of data throughout the entire process, providing a highly robust solution for information management in complex engineering environments.

[0035] like Figure 1As shown, this is the system's closed-loop feedback and dynamic optimization mechanism. When the visualization module detects that a user has rejected the repair suggestion signal generated by the verification module more than three times consecutively, it automatically generates a manual intervention request signal and transmits it to the project management terminal, triggering the engineering supervision party to intervene and review the dispute node. The manual intervention request signal includes a historical version comparison view of the disputed operation, the risk level evolution trend, and the cost impact analysis of the repair plan. After multi-factor authentication, the supervision party can choose to maintain the original suggestion, adopt the user's modification, or propose a third-party optimization plan in the arbitration interface. The decision result forms a manual arbitration signal and is fed back to the verification module. After receiving the arbitration signal, the verification module reassesses the risk level. If the arbitration plan introduces new conflict factors, it calls the lightweight simulation engine to perform multi-dimensional verification, such as simultaneously testing structural strength, pipeline flow, and fire evacuation routes, generating a new repair suggestion signal containing a verification report, and encrypting and storing the full log of the arbitration process in an independent audit branch of the blockchain evidence storage module. At the same time, the system establishes a dynamic learning mechanism. When there is a significant difference between the manual arbitration result and the AI ​​suggestion, it automatically extracts the features of the disputed case and adds them to the training set of the defect identification model. Through incremental learning, it optimizes the model's decision weights and reduces the probability of manual intervention in similar scenarios in the future. An environmental perception and supply chain collaboration module is introduced. This module connects to a meteorological data platform, a building material supplier database, and a construction machinery IoT system via API interfaces to achieve real-time fusion of external environmental and resource data. The environmental perception unit dynamically collects temperature, humidity, and wind speed data at the construction site and matches them with construction process parameters in the BIM model. For example, when continuous heavy rain is detected, the concrete pouring plan is automatically adjusted and linked to the model progress tag. At the same time, a construction plan change suggestion signal is generated and transmitted to the access management module, triggering a temporary upgrade of the permissions of relevant subcontractors to expedite the approval process. The supply chain collaboration unit monitors the inventory and logistics status of building materials in real time. When a delay in the delivery of key materials (such as prestressed steel strands) is detected, an alternative solution analysis process is automatically initiated. By comparing the parameters of alternative materials (such as tensile strength and elongation) in the supplier database, compliant alternatives are selected and a material replacement impact report is generated, including BIM model parameter adjustment suggestions, cost change estimates, and schedule rescheduling plans. This module also drives the blockchain evidence storage module to strengthen the evidence storage of supply chain change records. After the supplier qualification documents, logistics receipts and quality inspection reports are recognized by OCR, a structured evidence storage data package is generated. Each data package contains multiple layers of hash check codes to ensure the immutability of data throughout the entire supply chain.

[0036] Furthermore, a distributed knowledge graph is constructed to achieve global sharing of experience data. The knowledge transfer engine extracts historical BIM model version evolution records, risk management logs, and permission policy change trajectories from the blockchain storage module. It uses graph neural networks to extract common features (such as foundation design patterns under specific geological conditions and types of high-frequency construction conflicts) to generate reusable engineering knowledge units. Each knowledge unit contains three parts: first, the feature vector of the problem scenario (such as the selection parameters of the support scheme for deep foundation pits in soft soil); second, the association rules of the solution (such as the functional relationship between the spacing of support piles and soil parameters); and third, the quantitative evaluation index of the implementation effect (such as the percentage improvement in settlement control accuracy). When a new project imports the initial BIM model, the system automatically matches similar scenarios in the knowledge graph, pushes the associated knowledge units to the model building module, and drives the parametric modeling engine to pre-fill optimization parameters. For example, when a high-rise steel structure project is identified, the historical preferred design scheme of the wind vibration control node is automatically loaded. The multi-project collaboration framework allows different project teams to share anonymized knowledge units in an encrypted environment. Federated learning nodes perform differential privacy processing on the knowledge data contributed by each project, aggregate to generate a global optimization strategy, and back-update the defect identification model and rule engine of the local system, forming a positive cycle of cross-project experience. The framework also provides a smart contract-based knowledge trading market where enterprises can acquire expert-level solutions for specific scenarios by paying with digital tokens. All transaction records and knowledge usage traces are written to the consortium blockchain of the blockchain notarization module, ensuring clear intellectual property ownership and an automatically executed revenue-sharing mechanism.

[0037] The present invention provides a BIM-based engineering information management system that ensures data credibility through blockchain encryption and notarization, verifies model defects in real time with AI, improves collaboration efficiency through dynamic fusion of multi-source data, and achieves fine-grained access control through federated learning.

[0038] Therefore, the BIM-based engineering information management system of the present invention can solve the problems of data silos, security risks, low collaboration efficiency and insufficient intelligence in engineering information management.

[0039] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A BIM-based engineering information management system, characterized by, Comprise: a data processing module that receives raw engineering data and generates a structured data signal; a model building module that generates a BIM model from the structured data signal and extracts a model metadata signal and a dynamic operation log signal; the model building module is provided with a parametric modeling engine that generates a three-dimensional model skeleton from geometric parameters in the structured data signal and loads corresponding physical property calculation models through material attribute loading to generate geometric attribute signals and material parameter signals of the BIM model, while recording user editing operations on the model to form a dynamic operation log signal containing operation timestamp, user ID and modification content hash value, when detecting concurrent editing operations on the same model node, the model building module sends a version conflict warning signal to the blockchain storage module and triggers a version lock signal to suspend the editing authority of the current node; a blockchain storage module that encrypts and stores the metadata signal and generates a storage completion signal feedback to the model building module; the blockchain storage module uses an asymmetric encryption algorithm to perform sharding processing on the metadata signal, stores the sharded data blocks into private chain nodes and distributed file systems respectively, and writes the hash value of each data block into a blockchain account to form a storage chain, the version lock signal is realized through a smart contract, when more than three associated metadata blocks are continuously written in the storage chain, the smart contract automatically generates a version lock signal and transmits it to the model building module, the version lock signal is revoked by the blockchain storage module only after receiving the risk removal confirmation signal sent by the verification module; a verification module that analyzes the dynamic operation log signal and generates a risk warning signal and a repair suggestion signal when detecting abnormalities; the verification module is provided with a defect recognition model and a rule engine, the defect recognition model is trained through a historical engineering defect data set and is used to identify pipeline collision, structure overrun or specification conflict problems in the dynamic operation log signal, the rule engine loads an industry standard specification library and compares it with the BIM model attributes in real time, when detecting abnormalities, the verification module generates the risk warning signal containing risk level, positioning coordinates and repair priority label, and calls a preloaded simulation engine to simulate repair schemes to generate the repair suggestion signal containing parameter adjustment suggestions, alternative path planning and cost impact analysis, the repair suggestion signal triggers the data fusion module to start the model update process; a data fusion module that integrates the model metadata signal, the dynamic operation log signal and the repair suggestion signal to generate a fusion data signal, the fusion data signal is transmitted to the model building module and triggers the BIM model update; A permission management module adjusts user permissions according to the risk warning signal, generates an operation interception signal to a federal learning node to train an abnormal access model, and the federal learning node sends a permission policy update signal to the permission management module after training and synchronizes to the blockchain storage module to leave a trace.

2. The BIM-based engineering information management system of claim 1, wherein, The data processing module includes a data cleaning unit and a format conversion unit. The data cleaning unit filters redundant fields in the original engineering data through a preset regular expression template and a machine learning model, and generates an intermediate data signal after cleaning. The format conversion unit converts unstructured data into a unified coded structured data signal according to the BIM model construction standard after receiving the intermediate data signal, wherein the structured data signal includes geometric parameters, material properties and construction progress labels. The format conversion unit sends a data ready signal to the model construction module after completing the conversion, and synchronously encrypts the meta information of the structured data signal and transmits it to the blockchain storage module for pre-storage.

3. The BIM-based engineering information management system of claim 1, wherein, The data fusion module includes a space-time alignment unit and a conflict resolution unit. The space-time alignment unit normalizes the timestamps of IoT sensor data, unmanned aerial vehicle point cloud data and GIS data, and maps them to the corresponding positions of the BIM model through space coordinate system conversion to generate a space-time synchronized fusion data signal. When the sensor data deviates from the BIM model preset threshold by more than 10%, the conflict resolution unit automatically generates a data anomaly report and triggers a secondary verification of the verification module. The fusion data signal drives the model construction module to iteratively optimize the material parameters or structure topology of the BIM model, and sends real-time update instructions to the visualization module.

4. The BIM-based engineering information management system of claim 3, wherein, The permission management module includes a geofencing engine and a behavior analysis model. The geofencing engine generates a dynamic electronic fence according to the construction area boundary in the GIS data. When the user terminal GPS coordinates exceed the fence range, it automatically degrades the data access permission and generates an operation interception signal. The behavior analysis model obtains encrypted user operation history data through the federal learning node, trains an abnormal access detection model, and dynamically limits the user's editing permission for the key nodes of the BIM model when detecting irregular operation sequences. The interception log is encrypted and transmitted to the blockchain storage module.

5. The BIM-based engineering information management system of claim 1, wherein, The federal learning node aggregates multi-terminal user behavior data using homomorphic encryption technology, uploads model gradient parameters to the central server for aggregated update after training an abnormal access detection sub-model locally, generates a global permission policy update signal, and the permission policy update signal includes user role permission matrix, data sensitivity classification table and operation blacklist. The permission management module receives the update signal, starts secondary identity authentication for security risk users exceeding the threshold, and writes the updated strategy version hash value into the audit special chain of the blockchain storage module.

6. The BIM-based engineering information management system of claim 1, wherein, The model construction module further comprises a dynamic version merging unit which, after receiving the version locking signal sent by the blockchain storage module, automatically parses the concurrent modification records of the same BIM model node by multiple users, compares the differences in geometric properties, material parameters and associated logic in each version through a conflict resolution algorithm, and generates a version integration scheme containing weight scores and merging priority labels.

7. The BIM-based engineering information management system of claim 6, wherein, After the integration scheme is verified for structural stability by the lightweight simulation engine, the model construction module generates an updated BIM model version, and the merging log is encrypted and transmitted to the blockchain storage module to generate a new storage chain branch, and a version merging completion signal is sent to the permission management module to remove the editing operation restriction.

Images