A smart construction site construction management system based on BIM and Internet of Things

Abstract

The application discloses a kind of wisdom construction site construction management systems based on BIM and Internet of Things, it is related to building wisdom construction management technical field, in view of the problems in the prior art, the following scheme is presented, personnel, equipment, structure, material, environmental multi-source data are collected by Internet of Things global domain, data cleaning fusion is completed in edge end and BIM component real-time mapping is established;LSTM progress prediction model and risk grading evaluation model are constructed, and adaptive improved particle swarm optimization algorithm is used to solve multi-objective construction optimization scheme considering duration, cost, safety and environmental protection;The system is differentially managed and controlled in time period.This application solves the problems of traditional construction site BIM and Internet of Things data fragmentation, information silos, risk early warning lag, monitoring fault interruption control, etc., realizes construction site digital twin visualization, construction intelligent optimization scheduling, active early warning of safety hazards, automatic emergency disposal of faults, effectively reduces duration deviation, construction cost and safety accident rate.

Description

Technical Field

[0001] This invention relates to the field of intelligent construction management technology, and in particular to an intelligent construction site management system based on BIM and the Internet of Things. Background Technology

[0002] With the implementation of smart construction policies, BIM modeling and IoT sensing have been gradually applied to construction site management, but existing construction site management solutions have four major flaws: First, BIM and IoT data are disconnected, and digital models are out of touch with the actual physical conditions on site. Traditional BIM is only used for static 3D drawing display during the design phase. It does not connect with the real-time data collection of on-site sensors, equipment, and personnel. The model cannot be updated synchronously with dynamic information such as on-site personnel positioning, machinery operation, environmental temperature and humidity, structural stress, and foundation pit settlement. BIM is only used as a visualization tool and does not have real-time control capabilities, resulting in a lack of digital twin effect.

[0003] Second, there are data silos across multiple business areas, with progress, quality, safety, and materials belonging to separate independent systems; The software for real-name registration of construction site workers, tower crane monitoring, foundation pit monitoring, dust and noise monitoring, concrete temperature measurement, and construction progress monitoring is independent of each other, and the data protocols are not unified, making it impossible to achieve cross-module linkage analysis. For example, excessive foundation pit settlement cannot be automatically synchronized to BIM components and trigger a work stoppage warning. If the progress is delayed, it cannot be linked to material inventory to adjust the supply plan. It relies on manual cross-system data aggregation, resulting in delayed decision-making.

[0004] Third, risk warning is passive, and there is a lack of a multi-objective collaborative optimization scheduling mechanism; The existing system only has a single threshold alarm, which cannot be optimized in multiple dimensions such as construction period, material cost, safety risk and environmental constraints. High-risk areas, machinery overload and personnel crossing the boundary are only reminded after the fact. There is no closed-loop process for automatic task diversion, dynamic adjustment of process and automatic dispatch of hidden danger work orders. The coverage of manual inspection is limited, and it is difficult to cover the risks of hidden works and high-altitude operations in real time.

[0005] Fourth, there is a closed-loop iterative optimization and fault emergency migration mechanism; When changes in on-site working conditions, weather, or material delays cause prediction deviations, the system cannot automatically correct the BIM construction plan and early warning model; when large machinery or monitoring sensors malfunction, the monitoring data of the corresponding area is interrupted, and the system automatically migrates to the backup acquisition terminal when there is no monitoring task, which can easily create blind spots in management and control and lead to safety and quality accidents.

[0006] In summary, existing technologies are insufficient to achieve deep integration of BIM digital models and IoT site data, collaborative management of all elements, dynamic optimization scheduling, and emergency response to faults, and cannot meet the needs of refined, unmanned, and digital construction management for large-scale building projects. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a smart construction site management system based on BIM and the Internet of Things, which overcomes the deficiencies of existing technologies and effectively solves the problems in existing technologies.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A smart construction site management system based on BIM and IoT comprises a four-layer architecture: a site perception layer, an edge collaboration layer, a BIM cloud scheduling layer, and a mobile terminal application layer. Each layer communicates bidirectionally with encrypted MQTT / Modbus TCP. The site perception layer deploys multiple IoT data acquisition terminals to collect five categories of data: personnel, large equipment, structural environment, building materials, and construction plans. The edge collaboration layer is configured with an edge gateway, a collaboration controller, and a local data preprocessing module, and issues control commands via a CAN bus. The BIM cloud scheduling layer stores 3D BIM models with complete component attributes and is equipped with a global construction optimization engine. The mobile terminal application layer includes a central control screen, a management mobile app, and on-site tablets.

[0009] Preferably, the process includes the following steps: S1 full-domain data acquisition, S2 data preprocessing and BIM mapping, S3 multi-objective optimization control, S4 edge command issuance and differentiated management and control, S5 closed-loop feedback and BIM iterative update, S6 monitoring equipment fault linkage task migration, and S7 IoT monitoring equipment fault linkage migration management and control.

[0010] Preferably, the method steps S1-S7 are as follows: S1, Four-Tier Architecture Deployment and Comprehensive Multi-Dimensional Data Acquisition Establish a four-layer integrated smart construction site architecture: (1) On-site perception layer: Deploy IoT data acquisition terminals, including personnel positioning UWB tags, tower crane torque / tilt sensors, foundation pit settlement inclinometers, dust noise / temperature and humidity sensors, concrete temperature acquisition modules, video AI cameras, RFID material tags, tower crane black boxes, and electricity monitoring meters; sampling period is 100ms~200ms, and raw perception data is uploaded via Modbus TCP protocol.

[0011] The perception layer collects global data in five categories: ①Personnel data: Real-time coordinates of personnel, job type, certification status, and records of personnel crossing boundaries in high-risk areas; ② Large equipment data: tower crane lifting capacity, slewing angle, lifting height, equipment temperature, fault codes; ③Structural environmental data: foundation pit settlement, slope displacement, temperature difference between inside and outside concrete, PM10, noise, on-site temperature and humidity; ④ Building material data: quantity of steel bars / concrete delivered, remaining inventory, BIM code of components, and status of acceptance upon arrival; ⑤ Construction plan data: BIM model process duration, sub-project milestones, contract costs, and environmental control thresholds.

[0012] (2) Edge Collaboration Layer: Configures edge gateway, local data preprocessing module, lightweight BIM parsing unit, early warning judgment unit, improved particle swarm solver, and CAN bus command sending module; responsible for local real-time calculation to reduce cloud transmission pressure.

[0013] (3) BIM cloud scheduling layer: equipped with BIM 3D model library, big data storage and analysis platform, global construction optimization engine, model iteration training module, and fault operation and maintenance alarm module; stores complete building component BIM information (size, material, construction process, acceptance standard), realizes global strategy optimization and persistent data storage, and communicates bidirectionally with the edge collaboration layer through MQTT / HTTPS.

[0014] (4) Mobile terminal application layer: Management personnel use mobile APP, on-site tablets, and central control screen to realize BIM visualization, early warning reception, work order dispatch, progress adjustment, and fault reporting.

[0015] S2, Multi-source data preprocessing and dynamic mapping of BIM components in the edge collaboration layer The edge gateway receives heterogeneous data from the sensing layer and performs unified preprocessing: The 3σ criterion is used to remove abnormal sensor pulse data, and linear interpolation is used to complete the missing data due to network disconnection. This completes the time-series alignment and feature fusion of multi-source data. A unique coding mapping relationship is established between IoT-collected data and BIM components, binding personnel, equipment, and monitoring points to the corresponding spatial components of the BIM 3D model. The BIM model status is updated in real time, and a digital twin view of the construction site is constructed.

[0016] Construct two types of basic evaluation models: ① Construction progress prediction LSTM model: Input the time sequence data of personnel attendance, equipment operation and material arrival for one consecutive hour, and output the predicted progress values ​​of sub-items of the project for the next 1 / 3 / 6 days; ② Construction risk priority evaluation model, the evaluation indicators include the foundation pit deformation value, tower crane overload ratio, personnel crossing the boundary duration, concrete temperature difference, and dust exceeding the standard duration, weighted scoring formula: Weight Based on the scoring, high-risk queues, routine control queues, and low-risk delayed control queues are divided, with high-risk hazards being addressed first.

[0017] S3. Construction of Multi-Objective Construction Optimization Control Model With the optimization objectives of shortening the construction period, reducing overall construction costs, minimizing safety risks, and meeting environmental protection limits, a three-objective optimization function is established, taking into account factors such as the foundation pit deformation threshold, tower crane rated load, concrete curing temperature difference, dust emission constraints, and sub-project schedule constraints. (Construction period fluctuation rate): Reduce the deviation of the construction period in each process and avoid construction delays; (Comprehensive construction cost throughout the entire cycle): Optimize machinery scheduling and material delivery rhythm to reduce labor and equipment idle costs; (Construction site safety and environmental protection compliance rate): Minimize safety violations and dust and noise pollution issues.

[0018] S4. Adaptive Weight Improved Particle Swarm Optimization Algorithm for Solving Scheduling Schemes An improved particle swarm optimization algorithm is used to solve the multi-objective model, and adaptive inertia weights are introduced. , This represents the current iteration number. Maximum number of iterations; particle encoding includes tower crane operating time, personnel shift allocation, material arrival time, and monitoring equipment sampling frequency; fitness function integrates multiple constraints such as pit settlement limit, tower crane rated load, concrete curing temperature, and environmental emission threshold; solves and outputs global construction scheduling plan, equipment shift plan, and hazard classification and disposal strategy.

[0019] S5, Edge Gateway Command Issuance and Scenario-Based Differentiated Control Execution The edge collaboration layer translates the scheduling scheme into control commands and sends them to each IoT terminal: ① High-risk periods (foundation pit deformation approaching the threshold, tower crane operation in windy weather): Suspend high-risk processes, evacuate personnel from high-risk areas, and increase sensor sampling frequency to 100ms; ② During regular construction periods: Optimize the scheduling of tower cranes and work teams to balance the pace of material delivery; ③ Nighttime environmental control period: Reduce the operating power of dust-generating equipment, restrict high-noise machinery operations, and postpone construction to daytime.

[0020] Simultaneously generate hazard rectification work orders, which are pushed to on-site safety officers via mobile terminals, forming a rectification, review, and cancellation process.

[0021] S6. Global Data Closed-Loop Feedback and BIM Model Iterative Update Every 30 seconds, the edge layer compares the LSTM progress prediction value with the actual construction progress on site. When the progress deviation is greater than 10% or the risk monitoring data deviation is greater than 5%, the cloud scheduling layer retrieves the latest perception data to retrain the LSTM progress prediction model, simultaneously adjusts the weights of each objective function in the multi-objective optimization model, and re-solves the construction scheduling scheme; it also simultaneously updates the construction progress, acceptance status, and hidden danger records of BIM model components to achieve real-time synchronous iteration of the digital twin.

[0022] S7, IoT monitoring equipment fault linkage and migration control Real-time monitoring of sensor and data acquisition terminal operating status, and fault conditions are determined as follows: the equipment fails to upload data for 3 consecutive times, the acquired value exceeds the equipment's range by more than 2 times, and the equipment temperature is >65℃. Troubleshooting Procedures: ① Immediately mark the monitoring points corresponding to the BIM model as fault alarm status, and push the fault location and equipment type to the cloud and management personnel terminals; ② Search for backup IoT data acquisition terminals in the same area and calculate the coverage and sampling accuracy of the backup terminals; ③ Select the backup terminal with the highest coverage matching degree as the migration carrier, and send the monitoring task migration command through the CAN bus to switch the original faulty equipment monitoring point acquisition task to the backup terminal. ④ After the task migration is completed, re-execute the S3-S4 multi-objective optimization solution, update the risk warning strategy for the area, and eliminate monitoring and control blind spots.

[0023] According to claim 1, the smart construction site management system based on BIM and IoT is characterized in that the data collected by the site perception layer specifically includes: Personnel data: real-time 3D coordinates of workers, special operation certificate information, duration of stay in high-risk areas, and attendance records; Large equipment data: tower crane rated load, real-time lifting capacity, slewing angle, wire rope wear temperature, and shutdown fault codes; Structural monitoring data: cumulative settlement of the foundation pit, horizontal displacement of the slope, temperature difference between the inside and outside of the beam concrete, and stress of the reinforcing steel. Environmental data: On-site PM10 / PM2.5 concentration, noise level, site temperature and humidity, and sprinkler equipment operating status; Material BIM data: unique BIM code of components, quantity on site, stacking location, curing time, and acceptance status.

[0024] Preferably, the five types of global data collected by the field perception layer are as follows: Personnel data: 3D coordinates of workers, special operation certificate information, and duration of stay in high-risk areas; Large equipment data: Tower crane real-time lifting capacity, slewing angle, equipment casing temperature, and fault codes; Structural environmental data: foundation pit settlement, slope displacement, temperature difference between inside and outside concrete, dust, noise, site temperature and humidity; Building material data: unique BIM code for components, quantity received, stacking location, and curing time; Construction plan data: BIM model process duration, sub-project milestones, and environmental control thresholds.

[0025] Preferably, the weighted scoring formula for the risk prioritization assessment model is: Weight , For safety indicators, For quality indicators, For progress indicators, For environmental protection indicators.

[0026] Preferably, the improved particle swarm optimization algorithm speed update formula is as follows: Among them, learning factor The value is 2.0. The numbers are random numbers in the range of 0 to 1. The iteration termination condition is that the total change of the objective function is less than 0.01 for 5 consecutive generations.

[0027] Preferably, the nighttime environmental protection control period is from 22:00 to 6:00 the next day. The system automatically restricts the operation of high-noise machinery, reduces the power of dust-generating equipment, and postpones low-risk processes to daytime construction. The edge collaboration layer sends equipment control and monitoring task migration instructions to the Internet of Things terminal via the CAN bus.

[0028] Preferably, the mobile terminal application layer includes a central control screen, a management mobile APP, and an on-site construction tablet, used for BIM 3D visualization browsing, receiving early warning messages, processing rectification work orders, and adjusting construction plans.

[0029] Preferably, the LSTM construction progress prediction model takes into input one hour of continuous on-site time-series data and outputs the progress prediction results of the sub-items of the project for the next 1 day, 3 days, and 6 days.

[0030] Preferably, after the monitoring task migration is completed, the BIM cloud scheduling layer synchronously updates the status of the fault location and the regional risk warning threshold, and re-outputs the optimized construction scheduling plan. The edge collaboration layer completes local data preprocessing, warning judgment, and algorithm solution, and only summarizes the data and uploads it to the BIM cloud scheduling layer. The local command issuance delay is less than 200ms.

[0031] The beneficial effects of this invention are as follows: By deeply integrating BIM with the Internet of Things (IoT) to construct a real-time digital twin construction site, this invention establishes a four-layer hierarchical architecture and unifies the Modbus TCP / MQTT communication protocol. It enables the binding of on-site IoT measurement points with BIM component codes, and synchronizes dynamic data of personnel, equipment, structure, and environment to the 3D model in real time. This solves the problem of traditional BIM static display and disconnection from on-site working conditions, and realizes full-element visualized digital twin management and control of the construction site.

[0032] By integrating data from multiple business operations and eliminating information silos, multi-objective collaborative optimization is achieved. Data on personnel, machinery, structure, environmental protection, and materials are collected uniformly across all dimensions. Relying on an improved particle swarm optimization algorithm, the four major objectives of schedule, cost, safety, and environmental protection are balanced simultaneously. The system automatically outputs the optimal solutions for tower crane scheduling, material entry, and personnel allocation, eliminating the need for manual cross-system data aggregation and significantly improving the efficiency of construction decision-making.

[0033] The system proactively classifies and warns of risks, shifting management from post-event handling to pre-event prevention. It establishes a risk-weighted evaluation model, automatically classifies three levels of potential hazards, intensifies monitoring in high-risk areas, enforces work stoppages and personnel evacuations, and automatically generates rectification work orders for closed-loop circulation. It also relies on the LSTM progress prediction model to predict the risk of project delays, adjusts construction plans in advance, and significantly reduces the probability of safety accidents, rework, and project delays.

[0034] The system automatically migrates monitoring tasks in case of equipment failure, eliminating blind spots in control. When sensors or data acquisition terminals fail, the system automatically matches backup terminals and migrates monitoring tasks, and synchronously updates BIM fault point alarms to avoid interruption of key monitoring data such as foundation pits, tower cranes, and concrete, ensuring uninterrupted and refined control around the clock.

[0035] The system features a closed-loop iterative optimization across the entire domain, adapting to complex construction scenarios. It completes data comparison every 30 seconds, automatically retrains the prediction model and updates and optimizes weights when prediction deviations exceed the standard, continuously iterates construction scheduling strategies, and adapts to dynamic working conditions such as weather changes, material delays, and personnel movement. It is suitable for various building projects, including residential, municipal, and large public buildings.

[0036] Layered computing power collaboration reduces cloud pressure and provides fast response speed. The edge collaboration layer completes data preprocessing, local early warning, and algorithm solving, and only uploads the aggregated data to the cloud. The local command issuance latency is less than 200ms, enabling real-time response in high-risk scenarios and meeting the low-latency requirements of construction site safety management. The system was applied to a 32-story high-rise residential project (building area of ​​42,000 m2) and compared with the traditional manual management mode for 60 consecutive days: (1) the construction period fluctuation rate decreased from 38% to 12%, and the delay rate decreased by 63%; (2) the idle cost of tower cranes and labor decreased by 21%; (3) the early warning response time for safety hazards was shortened to 180ms, and the accident rate decreased by 68%; (4) the number of violations of dust and noise regulations decreased by 74%. The above data show that the system has achieved significant technical effects in four aspects: construction period control, cost optimization, safety early warning, and environmental compliance. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the overall four-layer architecture of the smart construction site of the present invention; Figure 2 This is a flowchart illustrating the migration process for monitoring equipment fault linkage tasks in this invention. Detailed Implementation

[0038] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0039] Example: Reference Figure 1-2 A smart construction site management system based on BIM and IoT, comprising a four-layer architecture, executing the complete S1~S7 control process.

[0040] S1, Four-Tier Architecture Deployment and Global Data Acquisition Establish a site perception layer, an edge collaboration layer, a BIM cloud scheduling layer, and a mobile terminal application layer.

[0041] The on-site perception layer deploys UWB positioning base stations, tower crane torque sensors, deep pit inclinometers, concrete embedded temperature measuring lines, online dust and noise monitoring instruments, AI PTZ cameras, RFID material tags, and power distribution cabinet meters on the construction site. A unified sampling period of 150ms is used to upload five types of data to the edge gateway via Modbus TCP. 1. Personnel data: Real-time 3D coordinates of construction personnel, special worker certifications, and duration of crossing boundaries in high-risk areas; 2. Tower crane equipment data: real-time lifting capacity, slewing angle, winch temperature, and fault codes; 3. Structural monitoring data: daily settlement of the foundation pit, slope displacement, and temperature difference between the inside and outside of the beam concrete; 4. Environmental data: PM10, noise, site temperature and humidity, sprinkler start / stop status; 5. Material BIM data: BIM codes, stacking locations, and quantities of steel bars, formwork, and precast components.

[0042] The edge collaboration layer deploys industrial edge gateways, integrating data preprocessing modules, lightweight BIM parsing units, improved particle swarm solvers, and local early warning judgment units; the BIM cloud scheduling layer deploys server clusters to store complete building BIM 3D models, including component dimensions, construction procedures, and acceptance standards for each floor; and the mobile terminals are configured with a project control screen, a mobile APP for project managers / safety officers, and on-site construction tablets.

[0043] S2, Multi-source data preprocessing and BIM dynamic mapping, model building After receiving heterogeneous sensing data, the edge gateway performs preprocessing: it uses the 3σ criterion to filter out instantaneous overload of tower cranes and abnormal values ​​of instantaneous dust pulses; it uses linear interpolation to fill in missing data due to network interruption; it completes time sequence alignment by unifying the time axis, extracts monitoring features and binds them with unique codes of BIM components, loads the BIM model on the central control screen, and automatically highlights the points exceeding the standard in red.

[0044] A short-term progress prediction model based on LSTM was trained using historical data collected continuously for one hour, outputting progress predictions for sub-projects for the next 1, 3, and 6 days; a risk-weighted evaluation model was constructed, with weights... Safety, quality, schedule, For environmental protection, the system automatically categorizes risks into three levels: high, normal, and low, prioritizing the handling of high-risk hazards.

[0045] S3. Multi-objective construction optimization model construction Three optimization objectives are set: minimize the project duration fluctuation rate, minimize the combined cost of manual and mechanical labor, and maximize the safety and environmental compliance rate; constraints include the tower crane rated load, foundation pit settlement threshold, concrete curing temperature difference, dust emission limit, and planned project duration, and corresponding mathematical objective functions are constructed.

[0046] S4. Solving with an adaptive improved particle swarm optimization algorithm The algorithm is set with a maximum of 100 iterations, a population size of 50, an adaptive inertia weight of 0.8~0.4, and a learning factor of [missing information]. The particle encoding includes the daily operating hours of the three tower cranes, the personnel allocation of the three construction teams, the arrival date of prefabricated components, and the sampling frequency of the monitoring equipment; the iteration termination condition is that the objective function changes by less than 0.01 for five consecutive generations; the solution outputs the optimal construction schedule, material arrival plan, and hierarchical monitoring strategy.

[0047] S5, edge command issuance and time-based differentiated control The edge gateway will convert the scheduling scheme into control commands and send them to each IoT terminal: 1. When the settlement of the foundation pit is close to the threshold, or during high-risk periods of strong winds (wind force ≥ level 6): suspend tower crane hoisting, evacuate personnel around the foundation pit, and increase sensor sampling frequency to 100ms; 2. Daytime routine construction (7:00-22:00): Optimize the scheduling of machinery and work teams to balance the pace of material delivery; 3. Nighttime environmental protection period (22:00-6:00 the next day): shut down high-noise cutting machinery, run the dust suppression spray equipment intermittently, and postpone low-risk processes to the daytime.

[0048] The system automatically generates electronic work orders for hazard rectification, which are then pushed to the safety officer's APP. The item can only be closed after the rectification is completed, photographed, and verified.

[0049] S6. Closed-loop feedback and BIM model iterative update The system retrieves real-time progress and monitoring data from the site every 30 seconds and compares it with the LSTM prediction value. When the deviation of the sub-item schedule is greater than 10% or the deviation of the foundation pit settlement monitoring is greater than 5%, the system retrieves the latest site data from the cloud to retrain the progress prediction model, adjusts the weight coefficients of the multi-objective optimization model in sync, and re-solves the scheduling scheme. The system also updates the construction progress, hidden danger records, and acceptance status of each component in the BIM model in sync, achieving real-time synchronization of the digital twin.

[0050] S7, Monitoring Equipment Fault Linkage Migration Control The edge gateway polls all IoT terminals for data transmission in real time. Fault determination conditions are: the terminal has no data for 3 consecutive times, the collected value exceeds twice the device's range, or the device casing temperature exceeds 65°C. Any condition triggered will result in a fault.

[0051] Troubleshooting Procedures: 1. Mark the corresponding monitoring points in the BIM model with red fault alarms and push the equipment number and fault location to the management personnel's APP; 2. Search for backup data acquisition terminals in the area of ​​the foundation pit / tower crane, compare their coverage and sampling accuracy, and select the terminal with the best matching degree; 3. Send a monitoring task migration command via CAN bus to switch the original fault location acquisition task to the backup terminal; 4. After the task migration is completed, re-execute the S3 and S4 multi-objective optimizations, update the risk warning threshold and monitoring frequency for the area, and eliminate blind spots in control.

[0052] Actual test results This system was applied to a smart construction site for a 32-story high-rise residential building, and compared with the traditional manual management mode in a 60-day continuous test. The following results were obtained: 1. The construction period fluctuation rate decreased from 38% to 12%, and the delay rate of sub-projects decreased by 63%; 2. Tower crane and labor idle costs decreased by 21%, and material inventory losses upon arrival were reduced by 17%; 3. The early warning response time for safety hazards in foundation pits, tower cranes, and high-altitude operations has been shortened to within 180ms, and the accident rate has decreased by 68%. 4. The number of environmental violations related to dust and noise pollution decreased by 74%, and the site successfully passed the civilized construction site inspection. 5. The task migration is completed within seconds after the monitoring equipment fails, with no interruption of critical monitoring data and no blind spots in control. 6. The workload of project management personnel in offline inspections is reduced by 55%, achieving lightweight, unmanned management and control of construction sites.

[0053] This system can be adapted to various construction projects such as high-rise residential buildings, commercial complexes, municipal roads, and bridges. It can be adapted to different engineering scenarios simply by adjusting the sensor deployment parameters, model weights, and equipment fault thresholds.

[0054] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to this embodiment. Any equivalent substitutions or minor optimizations made based on the technical concept of the present invention shall fall within the scope of protection of the present invention.

[0055] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A smart construction site management system based on BIM and the Internet of Things, characterized in that, The system comprises a four-layer architecture: a site perception layer, an edge collaboration layer, a BIM cloud scheduling layer, and a mobile terminal application layer. Each layer communicates bidirectionally with encrypted MQTT / Modbus TCP. The site perception layer deploys multiple IoT data acquisition terminals to collect five categories of data: personnel, large equipment, structural environment, building materials, and construction plans. The edge collaboration layer is configured with an edge gateway, a collaboration controller, and a local data preprocessing module, and issues control commands via a CAN bus. The BIM cloud scheduling layer stores 3D BIM models with complete component attributes and is equipped with a global construction optimization engine. The mobile terminal application layer includes a central control screen, a management mobile app, and on-site tablets.

2. The smart construction site management system based on BIM and IoT as described in claim 1, characterized in that, Includes the following steps: S1 Full-domain data acquisition, S2 Data preprocessing and BIM mapping, S3 Multi-objective optimization control, S4 Edge command issuance and differentiated management and control, S5 Closed-loop feedback and BIM iterative update, S6 Monitoring equipment fault linkage task migration, S7 IoT monitoring equipment fault linkage migration management and control.

3. The smart construction site management system based on BIM and IoT as described in claim 2, characterized in that, The specific steps of the methods S1-S7 are as follows: S1. The on-site perception layer deploys multiple types of IoT data acquisition terminals with a sampling period of 100ms~200ms. It collects five types of full-domain data: personnel, large equipment, structural environment, building materials, and construction plans, and uploads them to the edge collaboration layer via ModbusTCP. The BIM cloud scheduling layer stores three-dimensional BIM models with complete component attributes and transmits global strategies and model data bidirectionally with the edge collaboration layer via MQTT / HTTPS. S2. The edge collaboration layer preprocesses the collected data: it uses the 3σ criterion to remove pulse abnormal data, linear interpolation to complete missing data, completes time sequence alignment and feature fusion, and establishes the mapping relationship between IoT measurement points and BIM component codes; it constructs an LSTM construction progress prediction model and a risk priority evaluation model, and divides hidden dangers into high-risk queues, regular control queues, and low-risk delayed control queues based on weighted scores. S3. The multi-objective optimization control logic is as follows: the collaborative controller calls the pre-stored construction parameter database in the memory, verifies each constraint condition in order of priority, and iteratively solves the global optimal combination of equipment scheduling, personnel allocation, and material entry plan under the premise that the constraint conditions are met. S4. The model is solved using an improved particle swarm optimization algorithm with adaptive inertia weights. The formula for the adaptive inertia weights is as follows: The particle encoding is used for equipment scheduling, personnel allocation, and material arrival plans. The solution outputs a global construction scheduling plan. S5 and the edge collaboration layer will convert the scheduling scheme into control instructions and send them to each IoT terminal. Differentiated management and control will be implemented for high-risk construction periods, regular daytime periods, and nighttime environmental protection periods, and the closed-loop circulation of hazard rectification work orders will be automatically generated. S6. Every 30 seconds, compare the LSTM prediction value with the actual on-site working conditions. When the progress deviation is greater than 10% or the monitoring data deviation is greater than 5%, the cloud retrains the prediction model, updates and optimizes the model weights, and synchronously updates the construction status of the BIM model components. S7. Real-time determination of IoT terminal faults. The determination conditions are: no data upload for 3 consecutive times, value exceeding the range by 2 times, and equipment temperature > 65℃. After the fault, mark the BIM model point alarm, search and match the optimal backup acquisition terminal, issue instructions to migrate the monitoring task, and re-execute the multi-objective optimization scheduling after the task migration is completed.

4. The smart construction site management system based on BIM and IoT as described in claim 1, characterized in that, The five types of global data collected by the field perception layer are as follows: Personnel data: 3D coordinates of workers, special operation certificate information, and duration of stay in high-risk areas; Large equipment data: Tower crane real-time lifting capacity, slewing angle, equipment casing temperature, and fault codes; Structural environmental data: foundation pit settlement, slope displacement, temperature difference between inside and outside concrete, dust, noise, site temperature and humidity; Building material data: unique BIM code for components, quantity received, stacking location, and curing time; Construction plan data: BIM model process duration, sub-project milestones, and environmental control thresholds.

5. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, The weighted scoring formula for the risk priority assessment model is as follows: Weight , For safety indicators, For quality indicators, For progress indicators, For environmental protection indicators.

6. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, The improved particle swarm optimization algorithm speed update formula is as follows: Among them, learning factor The value is 2.

0. The numbers are random numbers in the range of 0 to 1. The iteration termination condition is that the total change of the objective function is less than 0.01 for 5 consecutive generations.

7. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, The nighttime environmental protection control period is from 22:00 to 6:00 the next day. The system automatically restricts the operation of high-noise machinery, reduces the power of dust-generating equipment, and postpones low-risk processes to daytime construction. The edge collaboration layer sends equipment control and monitoring task migration instructions to IoT terminals via the CAN bus.

8. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, The mobile terminal application layer includes a central control screen, a management mobile APP, and an on-site construction tablet, which are used for BIM 3D visualization browsing, receiving early warning messages, processing rectification work orders, and adjusting construction plans.

9. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, The LSTM construction progress prediction model takes into input one hour of continuous on-site time-series data and outputs the progress prediction results of sub-items of the project for the next 1 day, 3 days, and 6 days.

10. A smart construction site management system based on BIM and IoT as described in claim 3, characterized in that, After the monitoring task migration is completed, the BIM cloud scheduling layer synchronously updates the status of the fault location and the regional risk warning threshold, and re-outputs the optimized construction scheduling plan. The edge collaboration layer completes local data preprocessing, warning judgment, and algorithm solution, and only summarizes the data and uploads it to the BIM cloud scheduling layer. The local command issuance delay is less than 200ms.

Images