Monitoring information platform for intelligent identification and process management of dangerous and large engineering

Abstract

The invention relates to the technical field of supervision, in particular to a supervision information platform for intelligent identification and process management of dangerous and large projects. The system comprises a construction scene simulation unit, a dangerous large node marking unit, a multi-scene fusion comparison unit and an early warning data display unit. Through standardized modeling and flow demonstration, auxiliary construction plan making, resource allocation and progress management, real-time data and AI model combination is realized, risk points in the construction process are dynamically identified, early warning is carried out in advance, safety accidents are reduced, the identification accuracy and the automation level are improved, and meanwhile, multi-source monitoring data fusion and image comparison are combined, so that the construction efficiency is improved. According to the method, automatic deviation identification is achieved, the method can be used for project review, problem tracing and construction optimization based on historical stage models and early warning data, meanwhile, through intelligent pushing and dynamic display, supervision response efficiency and transparency are improved, and graphical display facilitates communication and cooperation among different professionals and different posts.

Description

Technical Field

[0001] The present invention relates to the field of supervision technology, and in particular to a supervision information platform for intelligent identification and process management of dangerous and major projects. Background Art

[0002] Dangerous and major projects (full name: highly dangerous sub-projects) refer to sub-projects that may cause mass casualties or major economic losses during the construction process, such as foundation pit projects, formwork projects, lifting and installation and disassembly projects, scaffolding projects, and demolition and blasting projects. During the construction of dangerous and major projects, combined with the trend of digital management, the process includes four stages: identification → planning → implementation → acceptance, among which: Identification stage: Before construction, the construction unit should analyze and mark the nodes of dangerous and major projects based on the design documents, geological survey reports, and other analytical drawings. Planning stage: The construction unit prepares a special construction plan, which is reviewed by the technical person in charge of the construction unit and signed by the chief supervisor engineer. Implementation stage: The project manager needs to give instructions to all operating personnel, conduct daily inspections, and the supervisor unit supervises on-site. Acceptance stage: Acceptance is carried out in stages, with joint signatures from the construction, supervision, and monitoring units, and the special plan, monitoring report, and acceptance records are compiled. Current supervision information platforms for intelligent identification and process-based management of hazardous and major projects rely heavily on the experience and inspections of operators. However, the accumulation of subtle deviations during the construction of hazardous and major projects can lead to subsequent uncontrollable situations. Human supervision is ineffective and has low precision. Deploying a large number of sensors for monitoring makes it inconvenient to view construction progress and differences in real time, making subsequent investigations time-consuming and labor-intensive. Therefore, we propose a supervision information platform for intelligent identification and process-based management of hazardous and major projects that simulates construction scenarios and compares monitoring data features to output construction progress and feature differences, and intelligently displays marked hazardous and major nodes. Summary of the Invention

[0003] The purpose of the present invention is to provide a supervision information platform for intelligent identification and process-based management of dangerous and major projects, so as to solve the following problems raised in the above background technology: Insufficient awareness of construction status: In traditional construction management, managers lack intuitive awareness of the dynamic changes in risks of critical projects, making it difficult to accurately predict risks based on drawings or schedules. Delayed identification of critical nodes: Risk points rely on manual experience to mark, which can be easily missed or misjudged. Furthermore, there is an inability to quantify priorities, leading to irrational resource allocation. Difficulty integrating multi-source data: Surveillance video, BIM models, and schedule data are isolated, making it impossible to compare deviations between actual construction and simulated scenarios in real time; Low early warning response efficiency: Traditional early warning systems rely on manual inspections or single sensors, resulting in information transmission delays and a lack of visual display, which affects decision-making efficiency. Uncontrollable personnel behavior: Unable to detect in real time the entry of unauthorized personnel or the violation of safety regulations by workers, leading to safety hazards; To achieve the above objectives, the present invention provides a supervision information platform for intelligent identification and process management of dangerous and major projects, including a construction scene simulation unit, a dangerous and major node marking unit, a multi-scene fusion comparison unit and an early warning data display unit; The construction scene simulation unit establishes a construction simulation scene based on the real scene of the dangerous and major project. The construction simulation scene includes a standardized 3D model and a staged visual demonstration diagram. The dangerous and major nodes are marked on the 3D model according to the construction stage through the dangerous and major node marking unit, and the dangerous and major nodes are displayed based on the hazard weight ranking. The multi-scene fusion comparison unit is used to integrate multi-scene monitoring data of dangerous and major projects in real time to form a VR image, divide the VR image into multiple feature images according to the area where the dangerous and major nodes are located, and output staged visual demonstration images mapped by the feature images in sequence according to the order, and calculate the similarity between the feature images and the staged visual demonstration images; The warning data display unit is used to preset a difference threshold, receive feature images with a similarity lower than the difference threshold, send a warning signal, and dynamically display the warning reasons of the dangerous nodes.

[0004] As a further improvement of this technical solution, the construction scene simulation unit includes a stage division module, a model building module and a dynamic demonstration module; The phase division module is used to divide the entire project into multiple construction phases using project management data and construction process knowledge. Each phase corresponds to a construction status, including structural status, equipment layout, personnel distribution and risk points. The model building module is used to build a standardized 3D model based on BIM technology, and generate a standard model for each stage output by the stage division module; The dynamic demonstration module is used to arrange the standardized 3D models of each stage according to the timeline to form a dynamic process demonstration diagram, supporting timeline playback, stage jump, multi-view switching, and animation simulation.

[0005] As a further improvement of this technical solution, the critical node marking unit includes a node marking module and a critical quantification ranking module; The node marking module is used to establish a risk factor database, extract the construction content of the 3D model of the current construction stage in the model building module, input the construction content into the risk factor database, and output the risk factors of all dangerous nodes in the 3D model of the current stage; The risk quantification and sorting module is used to preset the quantitative dimension and scoring rules of each risk factor to establish a risk factor scoring rule table, sequentially receive multiple risk factors output by the node marking module to the risk factor scoring rule table, output weighted scoring values ​​to define risk values, and use bubble sorting to sort multiple risk factors.

[0006] As a further improvement of the present technical solution, the dangerous node marking unit also includes a model highlighting display module, which is used to receive the risk factors of all dangerous nodes in the 3D model of the current stage of the node marking module, and receive the multiple risk factor rankings output by the risk quantification ranking module, set a color mapping relationship in which the higher the risk value, the darker the color, and highlight multiple risk factors in the 3D model according to the color mapping relationship.

[0007] As a further improvement of this technical solution, the multi-scene fusion comparison unit includes a multi-source data acquisition module, an actual highlight display module and a feature matching module; The multi-source data acquisition module is used to divide the final 3D model into multiple monitoring areas, assigning several cameras to each area so that the monitoring range of multiple cameras covers the entire monitoring area, and matching the camera installation positions in the 3D model with the actual locations of dangerous and major projects to obtain actual VR images; The actual highlight display module is used to calculate the actual risk factor of the 3D model risk factor in the VR image through the node marking module, and use the model highlight display module to highlight the actual risk factor; The feature matching module is used to sort the actual risk factors based on the hazard quantification sorting module, and use the image similarity algorithm to compare the actual risk factor features and the risk factor features in the 3D model in turn, and output a similarity value. If the similarity value is greater than the matching threshold, a normal signal is output; if the similarity value is not greater than the matching threshold, a construction warning signal is output.

[0008] As a further improvement of this technical solution, the multi-scene fusion comparison unit further includes a personnel movement line matching module, which is used to pre-collect standard personnel features of the staff, extract the current personnel features in the multi-source data acquisition module, and compare the matching degree between the current personnel features and the standard personnel features, including the following postures: Posture 1: If the output does not match, it is transmitted to the warning data display unit to issue an unrelated personnel warning signal; Posture 2. If the output matches, the features of the dynamic process demonstration diagram of the current personnel features arranged on the timeline are extracted based on the dynamic demonstration module, and defined as standard personnel movement line features. At the same time, the current personnel movement line features output by the multi-source data acquisition module on the timeline are extracted. The personnel movement line features include the path and work area features of each personnel feature. The matching degree of the standard personnel movement line features and the current personnel movement line features is compared. If the output matches, it is a normal signal. If the output does not match, it is transmitted to the warning data display unit to issue a work warning signal.

[0009] As a further improvement of this technical solution, the warning data display unit includes a stage-type warning module and a demonstration diagram calling module; The stage-based warning module is used to receive the construction warning signal from the feature matching module, identify the construction stage matched by the construction warning signal, and perform warning operations according to the construction stage; The demonstration diagram calling module is used to display two warning reason diagrams on split screen. The two warning reason diagrams include the actual risk factor VR image highlighted by the actual highlight display module in the current stage, and the dynamic process demonstration diagram output by the dynamic demonstration module before and after the time node of the current stage.

[0010] As a further improvement of this technical solution, the staged warning module includes a flashing light and a voice announcer; The number of the flashlights is consistent with the number of the multiple construction stages divided by the stage division module, and each construction stage is matched with a flashlight; The voice announcer is used to preset the reminder voice for each stage, identify the reminder voice that matches the stage of issuing the construction early warning signal, and emit the sound through the speaker.

[0011] As a further improvement of the present technical solution, the warning data display unit also includes a warning feedback module, which is used to establish a mapping relationship between the warning signal type and the feedback terminal. The warning signal types include construction warning signals, unrelated personnel warning signals and work warning signals. The warning signal type is identified and the warning cause is fed back to the matching terminal device via wireless communication.

[0012] Compared with the prior art, the present invention has the following beneficial effects: In the supervision information platform for intelligent identification and process management of dangerous and major projects, the construction scene simulation unit is used to simulate the construction scene, thereby realizing the visualization of the entire construction process, improving the management personnel's perception of the construction status, and can be used for new employee training and construction plan briefing to improve training efficiency and understanding. The dangerous and major nodes are marked on the 3D model according to the construction stage through the dangerous and major node marking unit, and are displayed based on the hazard weight ranking. This is conducive to assisting construction plan formulation, resource allocation, and progress management through standardized modeling and process demonstration, realizing the combination of real-time data and AI models, dynamically identifying risk points in the construction process, issuing early warnings, reducing safety accidents, and improving recognition accuracy and automation levels. At the same time, multi-source monitoring data fusion + image comparison can realize automatic deviation identification, and based on historical stage models and warning data, it can be used for project review, problem tracing and construction optimization. At the same time, through intelligent push + dynamic display, the efficiency and transparency of supervision response are improved, and graphical display facilitates communication and collaboration between personnel of different professions and positions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a block diagram of the overall structural principle of the present invention; Figure 2 This is a principle block diagram of the personnel movement matching module of the present invention.

[0014] The meaning of each number in the figure is: 100. Construction scene simulation unit; 200. Dangerous node marking unit; 300. Multi-scene fusion and comparison unit; 400. Early warning data display unit. DETAILED DESCRIPTION

[0015] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0016] Example 1 See also Figure 1-Figure 2 As shown, this embodiment provides a supervision information platform for intelligent identification and process management of dangerous and major projects, including a construction scene simulation unit 100, a dangerous and major node marking unit 200, a multi-scene fusion comparison unit 300 and an early warning data display unit 400; The construction scene simulation unit 100 establishes a construction simulation scene based on the actual scene of the dangerous and major project. The construction simulation scene includes a standardized 3D model and a staged visual demonstration diagram, which realizes the visual display of the entire construction process, improves the management personnel's perception of the construction status, and can be used for new employee training and construction plan briefing to improve training efficiency and understanding. The dangerous and major nodes are marked on the 3D model according to the construction stage through the dangerous and major node marking unit 200, and the dangerous and major nodes are displayed based on the hazard weight ranking. This is conducive to assisting construction plan formulation, resource allocation, and progress management through standardized modeling and process demonstration; To enhance visualization and immersion, the construction scene simulation unit 100 includes a stage division module, a model building module, and a dynamic demonstration module; The phase division module is used to divide the entire project into multiple construction phases using project management data and construction process knowledge. Each phase corresponds to a construction status, including structural status, equipment layout, personnel distribution and risk points. Project management data and construction process knowledge are input through data input, including the following data: engineering drawings (CAD, BIM), construction schedules (Project, Gantt charts), construction organization designs, safety specifications and standards, and real-time monitoring data (IoT sensors, video surveillance). Project files are parsed to extract process nodes and dependencies. Clustering algorithms (such as K-means) are used to group processes by job type and resources. Based on the clustering results and construction logic, several stages are divided (the clearer the stage details, the more detailed the construction status). Each stage is named and a construction description text is generated. The model building module is used to build standardized 3D models based on BIM technology. It generates a standard model for each stage output by the stage division module. The model includes: structural components (beams, columns, supports, etc.), equipment layout (cranes, elevators, scaffolding, etc.), personnel movement lines (simulated personnel paths, work areas, etc.), and safety signs and risk point markings. Specifically, it inputs CAD drawings and BIM models, and outputs standardized 3D model files (such as FBX and GLTF formats) based on the construction content description and standard component library (scaffolding, supports, cranes, etc.) of each stage; When building a standardized 3D model for each stage, components of the current stage are extracted from the BIM model, and the components are unified in terms of naming, material, color, etc. The standard equipment model library is called to automatically arrange equipment. Path planning algorithms (such as the A* algorithm) are used to simulate personnel movement routes. Risk areas (such as high-altitude work and cross-construction) are highlighted in the model based on the construction content and specifications. The dynamic demonstration module is used to arrange the standardized 3D models of each stage according to the timeline to form a dynamic process demonstration diagram. It supports timeline playback (simulating construction progress), stage jump (viewing models of specific stages), multi-perspective switching (top view, head-on view, section view, etc.), and animation simulation (such as lifting, demolition, concrete pouring, etc.). Specifically, it generates a timeline control according to the stage time, switches different stage models according to the timeline, and performs animation simulation on key actions (such as lifting, moving, and demolition). Pay attention to the use of keyframe interpolation algorithms to achieve smooth transitions. By supporting operations such as pause, play, jump, zoom, and perspective switching, users can easily perform interactive control.

[0017] Secondly, in order to dynamically identify risk points during the construction process, facilitate early warning in the actual project, and reduce safety accidents, the dangerous node marking unit 200 includes a node marking module and a dangerous quantitative ranking module; The node marking module is used to establish a risk factor database. It extracts the construction content (such as component composition, personnel movement lines, and construction stages) of the 3D model in the current construction phase from the model building module, inputs the construction content into the risk factor database, and outputs the risk factors of all critical nodes in the 3D model in the current phase. The following steps are involved in establishing a structured, scalable, and queryable risk factor database: Setting construction phases in Revit and exporting component lists for the corresponding phases. Revit API / Dynamo / IfcOpenShell can be used to extract component data. A risk factor database is designed to store risk factor rules and support rule engine calls. Construction content is matched to the risk factor database, and relevant risk factors are matched from the `risk_factors` table. For example, if the current phase is "main structure construction," factors such as "height work," "scaffolding erection," and "rebar tying" are matched. Components displayed in the 3D model are traversed through the Revit API (C# or Python). Spatial analysis and coordinate processing are performed on the 3D model to extract their location coordinates, and the area range of the critical nodes is highlighted. The hazard quantification and ranking module is used to preset the quantitative dimension and scoring rules of each risk factor to establish a risk factor scoring rule table. It sequentially receives multiple risk factors output by the node marking module to the risk factor scoring rule table, outputs weighted scoring values ​​to define the hazard value, and uses bubble sorting to sort multiple risk factors. Among them, the risk factor scoring rule table can determine the hazard value based on the quantitative dimensions of the components of the existing dangerous and major projects, for example:

[0018] The quantitative dimensions of risk factors and example values ​​are set as weighted rule scores, and the rule scores are used to represent the danger values, thus realizing digital supervision of dangerous and major projects.

[0019] In order to further improve the visualization and early warning effects, the dangerous node marking unit 200 also includes a model highlighting display module. The model highlighting display module is used to receive the risk factors of all dangerous nodes in the 3D model at the current stage from the node marking module, and receive the multiple risk factor rankings output by the risk quantification ranking module, and set a color mapping relationship in which the higher the risk value, the darker the color. In the 3D model, multiple risk factors are highlighted according to the color mapping relationship, which is conducive to visually focusing on risk factors through highlighting, and is conducive to staff focusing on them through simulation situations, and displaying risk value digital labels. At the same time, clicking on the node displays detailed scores.

[0020] Then, the multi-scene fusion and comparison unit 300 is used to integrate the multi-scene monitoring data of dangerous and major projects in real time to form a VR image, divide the VR image into multiple feature images according to the area where the dangerous and major nodes are located, and output the staged visual demonstration map of the feature image mapping in sequence according to the sorting, calculate the similarity between the feature image and the staged visual demonstration map, combine the real-time data with the AI ​​model, dynamically identify the risk points in the construction process, provide early warning, reduce safety accidents, and improve the recognition accuracy and automation level. At the same time, multi-source monitoring data fusion + image comparison can realize automatic deviation identification, and based on the historical stage model and warning data, it can be used for project review, problem tracing and construction optimization; Furthermore, the multi-scene fusion comparison unit 300 includes a multi-source data acquisition module, an actual highlight display module, and a feature matching module; The multi-source data acquisition module is used to divide the final 3D model into multiple monitoring areas. Each area is assigned several cameras, ensuring that the monitoring range of multiple cameras covers the entire monitoring area. The camera installation positions in the 3D model are matched to the actual locations of dangerous and major projects to obtain actual VR images. The "cone" model is used to estimate the camera coverage range. The field of view angle is usually 45° to 90°, and the installation height is recommended to be 3 to 8 meters (depending on the area size). The coverage radius is calculated based on the camera resolution and lens parameters. Tools such as SketchUp are used for visual point simulation to check whether the image stitching is complete. If the stitching is complete, the camera installation position is output. Otherwise, the simulation is repeated. The actual highlight display module is used to calculate the actual risk factor of the 3D model risk factor in the VR image through the node marking module, and use the model highlight display module to highlight the actual risk factor, which is conducive to directly viewing the dangerous nodes in the VR image corresponding to the actual monitoring video; The feature matching module is used to sort the actual risk factors based on the hazard quantification sorting module. It uses the image similarity algorithm to compare the actual risk factor features with the risk factor features in the 3D model and outputs a similarity value. If the similarity value is greater than the matching threshold, a normal signal is output. If the similarity value is not greater than the matching threshold, a construction warning signal is output. The image similarity algorithm essentially maps image data into a certain feature representation, and then calculates the differences or similarities between these features to obtain a similarity score (0-1 or 0-100%). The higher the similarity score, the more similar the two images are; the lower the score, the greater the difference. It involves extracting high-level semantic features of the image through a convolutional neural network (CNN) and then performing similarity calculations. Specifically, it uses a two-branch network with shared weights to extract features, compares the distance between feature vectors, and reflects similarity through distance.

[0021] And, as Figure 2 The multi-scene fusion comparison unit 300 also includes a personnel movement line matching module, which is used to pre-collect standard personnel features of the staff, extract the current personnel features from the multi-source data collection module, and compare the matching degree between the current personnel features and the standard personnel features, including the following postures: Posture 1: If the output does not match, it is transmitted to the warning data display unit 400 to issue an unauthorized personnel warning signal, so as to facilitate timely detection when unauthorized personnel enter the construction site of a dangerous project, thereby improving safety; Posture 2. If the output matches, the features of the dynamic process demonstration diagram of the current personnel features on the timeline are extracted based on the dynamic demonstration module, and defined as standard personnel movement line features. At the same time, the current personnel movement line features output by the multi-source data acquisition module on the timeline are extracted. The personnel movement line features include the path and work area features of each personnel feature. The matching degree of the standard personnel movement line features and the current personnel movement line features is compared. If the output matches, it is a normal signal. If the output does not match, it is transmitted to the early warning data display unit 400 to issue a work early warning signal to avoid workers from acting in disorder or not according to plan on the site of dangerous projects. On the one hand, it causes other workers or equipment to be unable to understand the situation and continue to work, causing dangerous accidents to the workers. On the other hand, it causes slack work that affects project efficiency.

[0022] The warning data display unit 400 is used to preset a difference threshold, receive feature images with a similarity below the difference threshold, issue a warning signal, and dynamically display the warning reasons for the critical nodes. Through intelligent push + dynamic display, the efficiency and transparency of the supervision response are improved, and the graphical display facilitates communication and collaboration between personnel of different professions and positions; Among them, the warning data display unit 400 includes a stage-by-stage warning module and a demonstration diagram calling module; The stage warning module is used to receive the construction warning signal from the feature matching module, identify the construction stage matched by the construction warning signal, and perform warning operations according to the construction stage; The demonstration diagram calling module is used to display two warning reason diagrams on split screen. The two warning reason diagrams include the actual risk factor VR image highlighted by the actual highlight display module in the current stage, and the dynamic process demonstration diagram output by the dynamic demonstration module before and after the current stage time node. Among them, before and after the current stage time node can be a dynamic process demonstration diagram displayed on the timeline formed by the previous stage → current stage → next stage, which is helpful for staff to see the difference problems.

[0023] It is worth mentioning that the staged warning module includes a flashing light and a voice announcer; The number of flashlights is consistent with the number of construction stages divided by the stage division module. Each construction stage is matched with a flashlight. The flashlight generates a pulse signal through a timing circuit (such as a 555 chip) or an MCU (single-chip microcomputer). The current passes through the LED to emit light, and the pulse signal controls the intermittent on and off of the LED to provide early warning. When the stage-based early warning module outputs the project early warning signal, the current construction stage can be determined based on the flashlight, providing a more accurate understanding of the on-site situation. The voice announcer is used to preset the reminder voice for each stage, identify the reminder voice that matches the stage of issuing the construction warning signal, and emit sound through the speaker. On the one hand, it reminds all on-site staff of the warning, and on the other hand, it further enables the staff to understand the on-site stage situation. Among them, the voice player outputs the voice signal through electronic synthesis or pre-recorded audio, and plays it through an external speaker.

[0024] The early warning data display unit 400 also includes an early warning feedback module, which is used to establish a mapping relationship between the early warning signal type and the feedback terminal. The early warning signal types include construction early warning signals, unrelated personnel early warning signals and work early warning signals. The early warning signal type is identified and the early warning reason is fed back to the matching terminal device through wireless communication. For example, when the early warning signal type is a construction early warning signal, the early warning reason is directly sent to the terminal (mobile phone APP, etc.) of the construction personnel at the current stage, so that each construction personnel can understand the early warning reason in time at the terminal, which is conducive to subsequent adjustments. Similarly, when the early warning is for unrelated personnel, the early warning reason is sent to the terminal of the staff responsible for maintaining the site. For work early warning signals, the early warning reason is sent to the staff terminal corresponding to the mismatch between the path of the personnel characteristics and the work area characteristics, so as to achieve accurate reminders and improve adjustment efficiency.

[0025] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A supervision information platform for intelligent identification and process-based management of hazardous and major projects, characterized by: It includes a construction scene simulation unit (100), a dangerous node marking unit (200), a multi-scene fusion comparison unit (300) and an early warning data display unit (400); The construction scene simulation unit (100) establishes a construction simulation scene based on the real scene of the dangerous and major project, the construction simulation scene including a standardized 3D model and a staged visual demonstration diagram, and marks the dangerous and major nodes on the 3D model according to the construction stage through the dangerous and major node marking unit (200), and displays the dangerous and major nodes based on the hazard weight ranking; The multi-scene fusion comparison unit (300) is used to fuse multi-scene monitoring data of dangerous and major projects in real time to form a VR image, divide the VR image into multiple feature images according to the area where the dangerous and major nodes are located, and output staged visual demonstration images mapped by the feature images in sequence according to the order, and calculate the similarity between the feature images and the staged visual demonstration images; The warning data display unit (400) is used to preset a difference threshold, receive a feature image with a similarity lower than the difference threshold, issue a warning signal, and dynamically display the warning reason of the dangerous node.

2. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 1 is characterized by: The construction scene simulation unit (100) comprises a stage division module, a model building module and a dynamic demonstration module; The phase division module is used to divide the entire project into multiple construction phases using project management data and construction process knowledge. Each phase corresponds to a construction status, including structural status, equipment layout, personnel distribution and risk points. The model building module is used to build a standardized 3D model based on BIM technology, and generate a standard model for each stage output by the stage division module; The dynamic demonstration module is used to arrange the standardized 3D models of each stage according to the timeline to form a dynamic process demonstration diagram, supporting timeline playback, stage jump, multi-view switching, and animation simulation.

3. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 2 is characterized by: The critical node marking unit (200) comprises a node marking module and a critical quantification ranking module; The node marking module is used to establish a risk factor database, extract the construction content of the 3D model of the current construction stage in the model building module, input the construction content into the risk factor database, and output the risk factors of all dangerous nodes in the 3D model of the current stage; The risk quantification and sorting module is used to preset the quantitative dimension and scoring rules of each risk factor to establish a risk factor scoring rule table, sequentially receive multiple risk factors output by the node marking module to the risk factor scoring rule table, output weighted scoring values ​​to define risk values, and use bubble sorting to sort multiple risk factors.

4. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 3 is characterized by: The critical node marking unit (200) further comprises a model highlighting display module, the model highlighting display module being used to receive the risk factors of all critical nodes in the 3D model at the current stage from the node marking module, and to receive the multiple risk factor rankings output by the risk quantification ranking module, setting a color mapping relationship in which the higher the risk value, the darker the color, and highlighting the multiple risk factors in the 3D model according to the color mapping relationship.

5. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 4 is characterized by: The multi-scene fusion comparison unit (300) includes a multi-source data acquisition module, an actual highlight display module and a feature matching module; The multi-source data acquisition module is used to divide the final 3D model into multiple monitoring areas, assigning several cameras to each area so that the monitoring range of multiple cameras covers the entire monitoring area, and matching the camera installation positions in the 3D model with the actual locations of dangerous and major projects to obtain actual VR images; The actual highlight display module is used to calculate the actual risk factor of the 3D model risk factor in the VR image through the node marking module, and use the model highlight display module to highlight the actual risk factor; The feature matching module is used to sort the actual risk factors based on the hazard quantification sorting module, and use the image similarity algorithm to compare the actual risk factor features and the risk factor features in the 3D model in turn, and output a similarity value. If the similarity value is greater than the matching threshold, a normal signal is output; if the similarity value is not greater than the matching threshold, a construction warning signal is output.

6. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 5 is characterized by: The multi-scene fusion comparison unit (300) further includes a personnel movement line matching module, which is used to pre-collect standard personnel features of the staff, extract the current personnel features in the multi-source data collection module, and compare the matching degree between the current personnel features and the standard personnel features, including the following postures: Posture 1: if the output does not match, it is transmitted to the warning data display unit (400) to issue an unrelated personnel warning signal; Posture 2. If the output matches, the features of the dynamic process demonstration diagram of the current personnel features arranged on the time axis are extracted based on the dynamic demonstration module, and defined as standard personnel movement line features. At the same time, the current personnel movement line features output by the multi-source data acquisition module on the time axis are extracted. The personnel movement line features include the path and work area features of each personnel feature. The matching degree of the standard personnel movement line features and the current personnel movement line features is compared. If the output matches, it is a normal signal. If the output does not match, it is transmitted to the warning data display unit (400) to issue a work warning signal.

7. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 6 is characterized by: The warning data display unit (400) includes a stage-by-stage warning module and a demonstration diagram calling module; The stage-based warning module is used to receive the construction warning signal from the feature matching module, identify the construction stage matched by the construction warning signal, and perform warning operations according to the construction stage; The demonstration diagram calling module is used to display two warning reason diagrams on split screen. The two warning reason diagrams include the actual risk factor VR image highlighted by the actual highlight display module in the current stage, and the dynamic process demonstration diagram output by the dynamic demonstration module before and after the time node of the current stage.

8. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 7 is characterized by: The staged warning module includes a flashing light and a voice announcer; The number of the flashlights is consistent with the number of the multiple construction stages divided by the stage division module, and each construction stage is matched with a flashlight; The voice announcer is used to preset the reminder voice for each stage, identify the reminder voice that matches the stage of issuing the construction early warning signal, and emit the sound through the speaker.

9. The supervision information platform for intelligent identification and process-based management of dangerous and major projects according to claim 8 is characterized by: The warning data display unit (400) also includes a warning feedback module, which is used to establish a mapping relationship between warning signal types and feedback terminals, the warning signal types including construction warning signals, irrelevant personnel warning signals and work warning signals, identify the warning signal type, and feed back the warning cause to the matching terminal device via wireless communication.

Images