Intelligent construction site management and control system based on BIM

Abstract

The invention relates to the technical field of building construction management, in particular to an intelligent construction site management and control system based on BIM (Building Information Modeling), which comprises a cloud service platform, at least one edge computing node and a plurality of site sensing and executing terminals, the field sensing and executing terminal is in communication connection with the edge computing node, and the edge computing node is in communication connection with the cloud service platform; the output end of the field data acquisition and preprocessing module and the output end of the local decision-making and real-time control module are connected with the input end of the communication management module. The output end of the communication management module is connected with the input end of the cloud service platform; the field sensing and execution terminal comprises a sensing terminal used for collecting environment state data and production element state data, and an execution terminal used for executing a control instruction. Through cooperation of cloud and edge calculation, deep fusion of a building construction site from a static BIM model to dynamic and real-time data is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of construction management, and particularly relates to a smart construction site management and control system based on BIM. BACKGROUND

[0002] With the transformation of the construction industry towards digitization and intelligence, the concept of smart construction site has emerged. In the prior art, the digital management of the construction site mainly relies on two types of systems: one is a project management system based on building information modeling (BIM), and the other is a site monitoring system based on Internet of Things (IoT).

[0003] The management system based on BIM usually creates a high-precision three-dimensional model in the design phase of the project, which is used for design coordination, collision detection and construction simulation (4D / 5D). However, the application of such a system in the construction phase has significant limitations: once the BIM model enters the construction phase, it often becomes a static "digital archive". Although the model contains rich geometric and attribute information, it lacks a mechanism for automatic data interaction with the real-time changes in the physical environment, equipment status and personnel activities of the construction site. The state of elements such as "people, machines, materials, methods and environment" in the construction site changes rapidly, and the traditional BIM model cannot automatically and real-time absorb these dynamic data, resulting in a serious disconnection between the model state on which management decisions are based and the actual state of the site. This disconnection between "static model" and "dynamic site" makes it difficult for BIM technology to play a real-time guidance and control value in the construction process, and its application is mostly limited to the pre-planning and post-archiving stages.

[0004] On the other hand, the monitoring system based on the Internet of Things collects site data by deploying various sensors (such as environmental monitoring sensors, tower crane safety monitoring sensors, personnel positioning tags and video monitoring cameras) at the construction site. Although this type of system can obtain real-time data, its architecture usually has inherent defects. Each monitoring subsystem (such as tower crane monitoring, dust and noise monitoring, video monitoring) is independently constructed and operated, and the data formats and communication protocols are not the same, forming a serious "data island". More importantly, its data processing mode is mostly "perception- upload-cloud judgment-alarm", that is, all raw data (especially high-bandwidth video stream data) are transmitted to the remote cloud platform for centralized processing without processing. The construction site environment is complex, and the network condition is unstable. This architecture not only puts a huge pressure on the network bandwidth, but also leads to a high delay from event occurrence to instruction generation, which cannot meet the demand of millisecond-level response in high real-time control scenarios such as personnel and machinery collision prevention.

[0005] In addition, the data analysis capability of the existing system is relatively simple. The early warning mechanism is mostly based on single-point and single threshold judgment (such as alarm when PM2.5 concentration exceeds the set value), and lacks the capability of fusion and deep mining of multi-source heterogeneous data (such as correlation analysis of personnel location, mechanical trajectory and video image). The one-sidedness of this analysis mode leads to the inability of the system to accurately identify complex and potential safety risks (such as personnel entering the blind area of the tower crane outside the working range). At the same time, the system generally lacks the capability of machine learning based on historical data and real-time data, and cannot perform predictive analysis on construction progress deviation, nor can it provide proactive warning and optimization scheduling scheme before safety hazards occur, and the decision support has serious lag.

[0006] Therefore, the existing technical architecture cannot effectively solve the core problems of deep fusion of static BIM model and dynamic construction site, real-time collaborative processing of massive multi-source data, and intelligent prediction and decision-making, which restricts the development of smart construction site technology in the direction of initiative and intelligence. It has become an urgent technical need in the field to develop a new type of management and control system that can break through data barriers, realize cloud-edge collaboration, and have intelligent decision-making capability. SUMMARY

[0007] Based on the above purpose, the present application provides a BIM-based smart construction site management and control system, comprising: a cloud service platform, at least one edge computing node, and a plurality of field sensing and execution terminals; the field sensing and execution terminals are in communication connection with the edge computing nodes, and the edge computing nodes are in communication connection with the cloud service platform; The cloud service platform comprises a BIM model integration and lightweight processing module, a multi-source data fusion and digital twin construction module, an artificial intelligence analysis and decision-making module, and a visual interaction and instruction issuing module; the output end of the BIM model integration and lightweight processing module is connected with the input end of the multi-source data fusion and digital twin construction module; The output end of the multi-source data fusion and digital twin construction module is connected with the input end of the artificial intelligence analysis and decision-making module; the output end of the artificial intelligence analysis and decision-making module is connected with the input end of the visual interaction and instruction issuing module; the output end of the visual interaction and instruction issuing module is connected with the input end of the edge computing node; The edge computing node comprises a field data acquisition and preprocessing module, a local decision and real-time control module, and a communication management module; the output end of the field data acquisition and preprocessing module is connected with the input end of the local decision and real-time control module; The output ends of the field data acquisition and preprocessing module and the local decision and real-time control module are connected with the input end of the communication management module; the output end of the communication management module is connected with the input end of the cloud service platform. The field perception and execution terminal includes a perception terminal for collecting environmental state data and production factor state data, and an execution terminal for executing control instructions.

[0008] Preferably, the BIM model integration and lightweight processing module includes a model analysis unit and a lightweight generation unit. The model analysis unit is used to analyze the file format of the original BIM model, extract the geometric information, attribute information and hierarchical relationship of the components, and identify the management and control sensitive attributes of the components in the construction stage based on a pre-defined construction management knowledge base, the management and control sensitive attributes including the construction process logic and safety management level to which the components belong. The lightweight generation unit is connected with the model analysis unit in data, and is used to dynamically determine the degree of model simplification according to the terminal performance requirements and application scenarios; the importance score of the components in the construction site is calculated, the importance score being determined based on the management and control sensitive attributes of the components, the spatial position of the components in the model and the connection relationship between the components and other components; the grid simplification algorithm is used to process the geometric information of the components according to the determined importance score and the degree of simplification, remove the internal invisible components and detailed features, while retaining the unique identifiers and key attributes of the components, to generate lightweight BIM models with different levels of details suitable for different terminals.

[0009] Preferably, the multi-source data fusion and digital twin construction module includes a data fusion unit and a twin update unit. The data fusion unit is used to establish a unified space-time reference, calculate the conversion parameters from the field coordinate system to the BIM model coordinate system by receiving the actual measurement coordinates of at least three reference points pre-measured in the construction site and the coordinates of the reference points in the BIM model, and using a coordinate conversion algorithm; for all received sensor data, the coordinates are unified according to the device installation positions and the conversion parameters; at the same time, the time stamps of all data sources are synchronized through the network time protocol, to realize the space-time alignment of multi-source data; The twin update unit is connected with the data fusion unit, and is used to update the space-time aligned data to the corresponding lightweight BIM model components according to the mapping relationship between the device ID and the components in the BIM model; when a data flow interruption or anomaly is monitored, the data values of the missing period are calculated through linear interpolation or state estimation algorithm according to the historical data variation law of the data source and other sensor data related to the data source, to ensure the continuity and integrity of the digital twin update.

[0010] Preferably, the artificial intelligence analysis and decision module comprises a progress prediction unit, a safety risk identification unit and a scheduling optimization unit. The progress prediction unit is configured to construct a dual-flow neural network model, wherein one flow inputs planned progress sequence data, and the other flow inputs real-time collected image recognition results, material arrival time sequence data and personnel work hour data; the correlation weight of each feature in the real-time data flow and the planned progress feature is calculated through an attention mechanism, the dual-flow features are fused to predict the progress completion at a future time point, and a delay process that has the greatest impact on a critical path is identified; The safety risk identification unit is configured to adopt a multi-modal fusion identification model to identify specific safety hazards while processing positioning data flow and video data flow; first, the accurate position of a target person or equipment in a digital twin and the type of the area to which the target person or equipment belongs are determined through positioning data, and the safety features of the corresponding area are extracted through a convolutional neural network analysis of the video images of the corresponding area; finally, based on a decision-level fusion algorithm, spatial position information and visual feature information are integrated to output safety risk identification results and confidence levels; The scheduling optimization unit is configured to establish an optimization model with the shortest duration, the lowest cost or the highest resource utilization rate as the target based on the progress prediction results and the safety risk identification results, and to solve the optimal personnel, mechanical and material scheduling scheme by using a heuristic algorithm.

[0011] Preferably, the field data acquisition and preprocessing module comprises a sensor network access unit, an image acquisition unit and a data preprocessing unit. The sensor network access unit is configured to support multiple industrial communication protocols, and to configure data acquisition frequencies and transmission formats that match different types of sensors; The image acquisition unit is configured to connect high-definition cameras and unmanned aerial vehicle devices, and to control the shooting angle, focal length and shooting interval thereof; The data preprocessing unit is connected with the sensor network access unit and the image acquisition unit, and is configured to perform data cleaning on the collected raw data, to eliminate abnormal values that are obviously beyond the physical range, and to smooth data fluctuations by using a sliding average filtering algorithm; for video stream data, a background subtraction method and a lightweight target detection algorithm are used to identify moving targets and their basic features, and to generate structured event description information; and the processed sensor data and the event description information are packaged to form a preprocessed data packet.

[0012] Preferably, the local decision and real-time control module comprises a rule engine unit and a real-time control unit. The rule engine unit is configured to store and manage a safety rule library issued by the cloud, wherein each rule contains a trigger condition, an execution action, a rule priority and a validity period; by continuously matching the preprocessed real-time data with the rule trigger conditions, the corresponding rule is triggered once a match is successful; The real-time control unit is connected with the rule engine unit, and is used for receiving an execution instruction of a triggered rule, generating a corresponding control signal according to content of the instruction, and directly driving a sound-light alarm, an access controller or a mechanical braking device on site to execute a corresponding action, so that a millisecond-level response is realized.

[0013] Preferably, the rule engine unit is further provided with a rule learning mechanism, when it is monitored that a locally triggered rule and a cloud-side issued decision instruction conflict for multiple times within a certain time, context data of the conflict event is automatically recorded, including environmental state, device state and operator information, and the data is uploaded to the cloud-side service platform as training data for rule optimization.

[0014] Preferably, the communication management module comprises a data uploading unit and an instruction receiving and distributing unit. The data uploading unit is used for compressing preprocessed data packets and local event data to be uploaded by using different compression algorithms according to data types and emergency levels; for data containing personal privacy information, differential privacy technology is used to add random noise of a corresponding strength for desensitization processing according to a privacy protection level evaluation result. The instruction receiving and distributing unit is used for receiving an instruction data packet from the cloud-side service platform, analyzing a target device address and execution content of the instruction, verifying integrity and safety of the instruction, and distributing the instruction to a local decision and real-time control module or a specified on-site execution terminal.

[0015] Preferably, the perception terminal comprises a personnel positioning tag, a mechanical running state sensor, an environmental monitoring sensor and an image acquisition device. The personnel positioning tag uses UWB technology to measure a tag signal arrival time through a positioning base station deployed on a construction site, and calculates two-dimensional or three-dimensional coordinates of the tag. The mechanical running state sensor comprises an inclination sensor, a weight sensor, an amplitude sensor and a wind speed sensor, and is used for acquiring real-time working parameters of a mechanical device. The environmental monitoring sensor comprises a particulate matter concentration sensor, a noise sensor and a temperature and humidity sensor, and is used for acquiring environmental data of a construction site. The image acquisition device comprises a fixedly installed high-definition spherical camera and a mobile deployed unmanned aerial vehicle device, and is used for acquiring on-site video stream data.

[0016] Preferably, the execution terminal comprises a sound-light alarm, an access gate and an intelligent handheld terminal. The sound-light alarm triggers sound of different frequencies and light of different colors according to a received control signal. The access gate controls a passage state according to a received on-off instruction. The intelligent handheld terminal is used for receiving and displaying scheduling instructions, construction drawings and safety warning information issued by the cloud, and uploading feedback information of on-site personnel to the edge computing node.

[0017] Advantages of the present application: 1. The present application realizes dynamic mapping of BIM models and real-time data on site through BIM model integration and lightweight processing modules, multi-source data fusion and digital twin construction modules. The geometric information and control sensitive attributes of components are associated with multi-source data such as sensor data, personnel location and mechanical operation state, so that the BIM model is no longer a static file, but a digital twin that reflects the state of the construction site in real time, thereby making up for the data disconnection problem of traditional BIM models in the construction phase.

[0018] 2. The system realizes data access, standardized processing and compression upload of different sensor networks through edge computing nodes and their communication management modules, and quickly issues cloud instructions to on-site terminals. The data preprocessing, real-time control and communication management of the edge node break down the data barriers of the originally dispersed monitoring subsystem, reduce the network bandwidth pressure, and realize millisecond-level response, meeting the high real-time safety control requirements.

[0019] 3. The artificial intelligence analysis and decision module realizes accurate identification of potential safety risks by comprehensively analyzing multi-source heterogeneous data such as personnel positioning, mechanical trajectory and video images through multi-modal fusion and deep learning technology. Through attention mechanism and decision-level fusion algorithm, the system can discover risk points in real time in the complex environment of the construction site, improve the accuracy and reliability of safety warning, and overcome the one-sidedness of traditional threshold warning.

[0020] 4. The present application realizes construction progress deviation prediction, critical path delay identification and optimal resource scheduling scheme generation by using historical data and real-time data to establish a double-flow neural network and a heuristic optimization model through the progress prediction unit and the scheduling optimization unit. The system can provide scientific construction scheduling suggestions and improve the initiative and decision-making efficiency of construction management before safety hazards occur.

[0021] 5. Through the cooperation of the local decision and real-time control module, the execution terminal and the perception terminal, the system can automatically execute safety rules, trigger alarms, control access control and mechanical equipment, and realize millisecond-level response. The rule learning mechanism can continuously optimize the control strategy, making the on-site management and control more intelligent and dynamic, and significantly improving the construction safety and management efficiency.

[0022] 6、Lightweight BIM model generation, multi-detail level processing, edge computing and cloud service collaboration ensure that the system can run efficiently in different terminals and different site environments, supporting large-scale multi-source data processing. Data desensitization and privacy protection technology ensures the safety of personnel privacy, further enhancing the widespread application and security of the system. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0024] Fig. 1 The figure is a schematic diagram of the framework of the system of the present application; Fig. 2 The figure is a schematic diagram of the framework of the sensing terminal of the present application; Fig. 3 The figure is a schematic diagram of the framework of the execution terminal of the present application. DETAILED DESCRIPTION

[0025] The present application will be described in detail below in combination with the drawings and specific embodiments. It should be noted here that, in order to make the embodiments more detailed, the following embodiments are the best, preferred embodiments, and other alternative ways can also be used by those skilled in the art to implement them; and the drawings are only used to more specifically describe the embodiments, and are not intended to specifically limit the present application.

[0026] Please refer to Figs. 1-3 The embodiment of the present application provides a smart construction site management and control system based on BIM. The cloud service platform is the core of the entire system, mainly including four modules: The main function of the BIM model integration and lightweight processing module is to integrate the three-dimensional geometric information in the BIM model and perform lightweight processing to adapt to the fusion of real-time data on site. Model information includes architectural design drawings, component information, construction progress, etc., but these information needs to be updated synchronously with real-time data in the actual construction phase. This module optimizes the algorithm so that the BIM model can adapt to the dynamic changes of the site data.

[0027] The multi-source data fusion and digital twin construction module fuses data from different sensors (such as environmental monitoring sensors, personnel positioning sensors, mechanical equipment state sensors, etc.) and constructs a digital twin. The digital twin is a virtual model reflecting the real-time state of the construction site, which can display the site environment and construction progress synchronously. This process makes the BIM model no longer static, but a dynamic, real-time updating system.

[0028] The artificial intelligence analysis and decision module is responsible for intelligent analysis of the fused multi-source data, identifying potential risks and problems such as equipment failure, personnel entering dangerous areas, etc. At the same time, the module uses machine learning and deep learning algorithms to extract rules from historical data and real-time data, conducts predictive analysis and optimization scheduling, and provides the basis for decision-making.

[0029] The visual interaction and instruction issuing module displays real-time data and analysis results through the user interface, providing visual construction site conditions and risk warnings. At the same time, it can issue decision-making instructions to edge computing nodes, and the instructions are executed through on-site sensing and execution terminals to ensure that decisions can be quickly implemented.

[0030] The edge computing node reduces the burden of cloud data transmission and improves system response speed through on-site data collection, preprocessing, local decision-making, and real-time control. The edge node includes the following modules: The on-site data collection and preprocessing module collects various data from the construction site through on-site sensor devices, including environmental data, mechanical operating status, personnel location, etc. The data is preliminarily processed, such as denoising and standardization, to facilitate more effective transmission and analysis.

[0031] The local decision-making and real-time control module makes local real-time decisions based on pre-set rules and the output of the artificial intelligence analysis module, such as adjusting equipment operating mode, issuing safety warnings, etc. This module ensures the rapid response of on-site management and avoids high latency caused by relying on cloud processing.

[0032] The communication management module is responsible for data communication and protocol conversion between the entire edge computing node and the cloud platform, on-site sensing and execution terminals, ensuring accurate data transmission and supporting real-time updates.

[0033] The on-site sensing and execution terminal realizes the sensing and execution functions of the site through collection and control. The sensing terminal collects environmental state data and production factor state data, and the execution terminal controls equipment, personnel or other site elements according to instructions.

[0034] The invention realizes the deep fusion of the construction site from static BIM model to dynamic and real-time data through the cooperation of cloud and edge computing. Real-time data collection and preprocessing of on-site sensors, intelligent analysis and decision-making support enable management personnel to make quick responses and decisions during the construction process. Through the real-time nature of edge computing, the problem of slow response caused by cloud delay is avoided, ensuring the safety and efficiency of the construction site. In addition, intelligent prediction and scheduling optimization functions effectively reduce resource waste and delays in construction, improving the controllability and accuracy of construction progress.

[0035] In one possible implementation, first, the system receives an original BIM model file, which can be in Revit, IFC or other standard BIM format. The model parsing unit identifies and parses the file format, and extracts various components in the model.

[0036] The geometric information (such as size, shape, spatial position) of each component, the attribute information (such as material, construction stage, construction unit) and the hierarchical relationship (such as the parent-child, adjacent or connection relationship between components) are extracted.

[0037] In combination with a pre-defined construction management knowledge base, the control-sensitive attributes of each component in the construction stage are identified. The sensitive attributes include the criticality of the component in the construction process, the construction sequence requirement, and the safety management level (such as high-altitude operation, load-bearing component, etc.). These information will be used for subsequent model lightening and construction management decision.

[0038] According to the performance requirements of different terminals (such as mobile terminal, tablet terminal or large monitoring terminal) and application scenarios (real-time visualization, construction monitoring, command and dispatch), the degree of model simplification is dynamically set.

[0039] The importance score of each component is determined by its control-sensitive attributes, spatial position and connection relationship with other components. For example, load-bearing structure, key node or component associated with multiple components have higher scores; while auxiliary components or internal invisible parts have lower scores.

[0040] The geometric information is processed to remove internal invisible components and complex details, while retaining the unique identifier and key attribute information of each component for subsequent system management, data query and decision analysis.

[0041] Lightweight BIM models with different levels of detail are generated to adapt to the display and computing capacity of different terminals, realizing multi-terminal collaborative work and visualization requirements.

[0042] The present application realizes the intelligent transformation of BIM model from static design to dynamic, controllable and multi-terminal adaptive in the construction stage, improving the practicality and efficiency of the intelligent construction site management system.

[0043] In one possible implementation, to ensure that the real-time data of the construction site and the BIM model have a consistent coordinate system, at least three reference points need to be preset on the construction site and accurately measured to obtain the actual coordinates of these reference points. At the same time, the coordinates of these reference points are also defined in the BIM model.

[0044] A coordinate conversion algorithm is used to calculate the conversion parameters between the field coordinate system and the BIM model coordinate system. In this way, the coordinate data measured on site can be accurately mapped to the corresponding position in the BIM model.

[0045] For all the data collected by sensors on site, the coordinate information is unified into the coordinate system of the BIM model according to the installation location of the sensor device and the conversion parameters, ensuring that all sensor data is consistent with the spatial information of the BIM model.

[0046] Through the Network Time Protocol (NTP), time synchronization of different data sources is achieved, ensuring that data from different devices and sensors has a unified timestamp, which allows for accurate comparison and analysis of changes in multi-source data.

[0047] The twin update unit maps the data that has completed spatio-temporal alignment to the components in the BIM model through device ID, accurately updating real-time data to the corresponding BIM components (such as building components, mechanical equipment, etc.). For example, if a sensor collects temperature, humidity, or location information, these data will be updated to the corresponding components in the model.

[0048] During actual construction, due to equipment failure, network problems, etc., data flow may be interrupted or abnormal. At this time, the twin update unit analyzes the historical data variation of the data source and the data of other sensors related to it, and uses linear interpolation or state estimation algorithm to calculate the missing data, ensuring that the update of the digital twin will not be interrupted due to data loss.

[0049] Through precise spatio-temporal data fusion and flexible digital twin update methods, the application effect of BIM technology in smart construction sites is improved, with strong real-time, accuracy and fault tolerance, greatly optimizing data management and decision support capabilities during construction.

[0050] In one possible implementation, the progress prediction unit integrates two input data streams by constructing a double-flow neural network model. The first stream inputs planned progress sequence data, which represents the original construction plan and the scheduled completion time of each task. The second stream inputs real-time data, including image recognition results on the construction site, material arrival time sequence data, and personnel work hour data.

[0051] The system calculates the correlation weight between each feature in the real-time data stream and the planned progress feature through an attention mechanism. Through this method, the model can find the most critical factors among multiple real-time input data, thereby weighting the progress completion at the future time point.

[0052] At the same time of progress prediction, the system can also identify the most delayed process that has the greatest impact on the critical path. The critical path refers to the most important task that affects the total project duration, and early identification of delayed processes helps to adjust the plan in time and reduce the delay in the project duration.

[0053] The safety risk identification unit combines two data streams - positioning data stream and video data stream. First, the system accurately determines the location of the target personnel or equipment in the digital twin and its area type through positioning data, ensuring real-time monitoring of all critical resources.

[0054] Images in the video data stream are input into a convolutional neural network (CNN) for analysis, extracting safety features such as unsafe personnel behavior or potential accident areas.

[0055] Based on the decision-level fusion algorithm, the system combines spatial location information and visual features to generate comprehensive safety risk identification results. Finally, the system outputs the identification results of safety risks with confidence values, helping construction managers accurately assess risk levels and potential threats.

[0056] The scheduling optimization unit establishes an optimization model based on the progress prediction results and safety risk identification results, aiming for the shortest duration, lowest cost, or highest resource utilization. The optimization model considers all possible constraints such as resource availability, construction progress, budget, etc.

[0057] A heuristic algorithm is used to solve the optimization model to obtain the optimal personnel, machinery, and material scheduling scheme. This algorithm can provide efficient scheduling solutions while considering various complex constraints, ensuring that construction projects can be completed on time and costs are effectively controlled.

[0058] Through the cooperation of these three modules, the system realizes comprehensive intelligent analysis and decision support for the construction site, effectively improving the accuracy and efficiency of construction management and promoting the construction of smart construction sites.

[0059] In one possible implementation, multiple industrial communication protocols are supported to connect with various industrial sensor devices, and different industrial communication protocols are compatible. These sensors can include temperature, humidity, pressure sensors, and location sensors, etc., for real-time data collection on the construction site.

[0060] The configuration of data collection frequency and transmission format can be flexibly configured according to the characteristics of different types of sensors to ensure efficient and accurate data collection and transmission for each type of sensor. Depending on the type of sensor, the collection frequency may vary to meet different data accuracy requirements.

[0061] The image acquisition unit can connect with high-definition cameras and unmanned aerial vehicle devices. Through these two devices, high-definition image data of the construction site can be obtained in real time.

[0062] The system automatically adjusts the shooting angle and focal length of the camera and the drone to ensure that the captured image information is accurate and clear. At the same time, the system automatically sets appropriate shooting intervals to ensure the timeliness and continuity of video image data for subsequent processing and analysis.

[0063] The data preprocessing unit first cleans the raw data from the sensors and image acquisition unit. For the collected sensor data, the system eliminates outliers that are obviously beyond the physical range, ensuring the authenticity and accuracy of the data.

[0064] When processing sensor data, the system uses a sliding average filtering algorithm that helps smooth out fluctuations in the data, reducing the impact of noise and making the data more stable for subsequent analysis.

[0065] For video stream data, the system uses background subtraction and lightweight target detection algorithms. Background subtraction can effectively extract foreground objects from video streams and remove background interference. The target detection algorithm helps identify moving targets in the video and extract their basic features, such as target type, location, speed, and other information. Through these technologies, the system can monitor and identify key dynamics in the construction site in real time.

[0066] After processing, the system converts the identified moving targets and their basic features into structured event description information. For example, the movement path of a certain device in a specific area, the abnormal behavior of a certain construction worker, and other information will be recorded for subsequent decision-making.

[0067] In one possible implementation, the rule engine unit first stores and manages the security rule library issued by the cloud. These rules are formulated by the cloud based on global monitoring data and historical analysis results, and include multiple elements such as trigger conditions, execution actions, rule priorities, and validity periods. Each rule has clear conditions and execution logic.

[0068] The rule engine continuously analyzes real-time data collected on site and matches it with trigger conditions in the rule library. When real-time data meet the trigger conditions of a certain rule, the rule engine will automatically trigger the corresponding rule and generate an execution instruction. These trigger conditions may be abnormal behavior of a certain device, dangerous signals in the construction environment, or violations of personnel, etc.

[0069] When the rule engine matches a certain rule based on preprocessed data, the execution instruction is sent to the real-time control unit. The main task of this unit is to receive these instructions and generate corresponding control signals based on the instruction content.

[0070] The real-time control unit automatically generates control signals based on the execution instructions issued by the rule engine. These control signals can directly drive various devices on site, such as sound and light alarms, access control controllers, or mechanical braking devices, etc.

[0071] Once the control signals are generated, the real-time control unit will immediately drive the corresponding on-site devices to perform the predetermined actions, ensuring that in the event of an abnormal situation, it can respond quickly within milliseconds, thereby timely eliminating potential safety risks.

[0072] The design and implementation of the local decision-making and real-time control module provide a real-time, efficient, and intelligent management platform, which can fine-tune the management of the construction site and ensure that safety hazards can be discovered and handled in a timely manner, thereby achieving the goal of a smart construction site.

[0073] In one possible implementation, when the rule engine unit detects that the locally triggered rules and the cloud-deployed decision instructions conflict multiple times within the same time period during execution, the system will start the conflict recording mechanism. These conflicts may be due to incomplete rule definitions or execution errors caused by significant changes in the on-site environment.

[0074] For each conflict event that occurs, the rule engine automatically records detailed context data. This includes: Environmental status: such as temperature, humidity, wind speed, safety level of the construction area, and all parameters related to the on-site environment.

[0075] Device status: including the status of devices related to rule triggering, such as the working status, fault status, and maintenance status of the device.

[0076] Operator information: including the operator's identity, operation record, operation time, and other information to ensure that human factors can be traced.

[0077] When the conflict event is recorded, the rule engine automatically uploads these context data to the cloud service platform. The data upload not only includes the original record, but also includes the detailed context of the conflict, providing a basis for subsequent rule optimization.

[0078] After the cloud platform receives this data, it combines historical data and existing rules to optimize and adjust the rules through machine learning algorithms. These uploaded conflict events and their context data serve as training data to help the cloud further optimize decision instructions and rule matching logic, thereby reducing the occurrence of future conflicts.

[0079] Through these steps and mechanisms, the rule learning mechanism not only improves the intelligence and flexibility of construction management, but also ensures that the system can continuously optimize according to changes in the actual environment to cope with more complex construction management challenges in the future.

[0080] In one possible implementation, the data upload unit classifies the data to be uploaded according to its type (such as construction environment data, equipment status data, personnel information, etc.) and urgency (such as equipment failure, construction accident, routine monitoring, etc.). Based on these evaluation results, a suitable compression algorithm is selected for data compression. For high-urgency or large-data-volume situations, efficient compression algorithms such as LZ77, Huffman encoding, etc. are used to ensure that data transmission time and bandwidth occupancy are reduced during uploading.

[0081] For data containing personnel privacy information (such as operator identity, location data, etc.), in order to ensure privacy security, the data upload unit will evaluate the privacy protection level and process the data using differential privacy technology. Differential privacy technology adds appropriate random noise to the data to ensure that personal privacy information cannot be restored by data when the data is published or stored, thereby effectively protecting the privacy of construction site personnel.

[0082] After compression and privacy processing, the data upload unit transmits the data to the cloud service platform to ensure data security and efficiency, facilitating subsequent analysis and decision-making.

[0083] The instruction receiving and distribution unit is responsible for receiving instruction data packets from the cloud service platform. These instructions may involve controlling field devices, adjusting construction parameters, or issuing safety warnings, etc.

[0084] After receiving the instruction data packet, the instruction receiving and distribution unit will parse the data packet to extract the address information of the target device and the execution content. The instruction content includes the operations to be performed (such as starting the device, adjusting the settings, etc.), while the address of the target device indicates the specific device or execution terminal that the instruction acts on.

[0085] Before passing the instruction to the field execution device, the instruction receiving and distribution unit will verify its integrity and security. By checking the signature, checksum, or timestamp of the data packet, it ensures that the instruction has not been tampered with and confirms its legal source. This step ensures that the instruction is not at risk of external attacks or data leakage during transmission.

[0086] After verification, the instruction will be sent to the local decision-making and real-time control module or other designated field execution terminal. The execution terminal receives the instruction and immediately performs the corresponding operation as required.

[0087] The communication management module improves the real-time performance and security of the system through efficient data upload processing and safe instruction distribution, optimizes resource management and decision execution of the construction site, and ultimately realizes efficient, intelligent, and safe management and control of the smart construction site system.

[0088] In one possible implementation, personnel positioning tags use UWB technology for positioning. UWB is a high-bandwidth, low-power wireless communication technology that can provide high-precision real-time positioning. Personnel positioning tags in the sensing terminal communicate with multiple positioning base stations deployed on the construction site through the built-in wireless transmission module. The positioning base stations calculate the two-dimensional or three-dimensional coordinates of the personnel tags by measuring the time of arrival (TOA) of the signals according to the triangulation principle. This positioning method has high precision (generally with an accuracy of a few centimeters), can monitor personnel positions in real time, and ensures the safety of personnel and the accuracy of positioning, especially in complex construction environments.

[0089] Mechanical operating state sensors are mainly used to monitor the working conditions of construction machinery. They include: Inclination sensor: monitors the inclination angle of mechanical equipment to prevent equipment failure or accidents caused by excessive inclination.

[0090] Weight sensor: real-time detection of equipment load conditions to ensure equipment works within the rated load range and avoid overloading.

[0091] Amplitude sensor: used to monitor the vibration amplitude of mechanical equipment to detect potential mechanical failures in a timely manner.

[0092] Wind speed sensor: detects wind speed, especially for overhead working equipment such as cranes and hoists, to ensure that the wind speed does not exceed the safe operating range.

[0093] Real-time data collection from these sensors provides a reliable basis for the maintenance and safety management of mechanical equipment on the construction site, helping to detect potential equipment failures and perform preventive maintenance in a timely manner.

[0094] Environmental monitoring sensors are used to collect environmental data on the construction site, mainly including: Particulate matter concentration sensor: used to monitor the concentration of dust particles in the air to ensure that the air quality on the construction site meets health standards.

[0095] Noise sensor: used to monitor the noise level on the construction site, especially the noise generated by construction machinery and working environment, to avoid excessive noise affecting the surrounding environment and workers' health.

[0096] Temperature and humidity sensor: monitors the temperature and humidity of the construction site environment to help adjust construction plans and provide necessary safety warnings in extreme weather conditions.

[0097] Environmental sensors help improve the construction environment, improve worker health protection, and meet relevant environmental protection regulations by monitoring air quality, noise, and weather conditions in real time.

[0098] Image acquisition equipment includes: Fixed high-definition ball camera: These cameras are usually installed in key locations (such as construction sites, entrances and exits, and near important equipment) for all-around monitoring of the construction site dynamics, especially for safety monitoring and progress tracking.

[0099] Mobile deployment of unmanned aerial vehicle equipment: Unmanned aerial vehicles can fly over the construction site as needed to capture real-time high-definition images and video stream data of the construction site, especially in difficult-to-reach areas such as high-altitude work and complex terrain.

[0100] Through image acquisition equipment, visual monitoring of the construction site can be achieved, which not only helps to monitor safety hazards and construction progress in real time, but also provides key evidence and emergency response support in emergency situations.

[0101] The various technologies of the perception terminal provide strong technical support for the safety, efficiency, and environmental protection of smart construction sites, making the management and control of construction sites more intelligent and automated, and ultimately promoting the entire construction industry to develop more efficiently and safely.

[0102] In one possible implementation, the function of the audible and visual alarm is to trigger different frequencies of sound and different colors of light according to the received control signal to issue a warning. In the system, the audible and visual alarm receives data from the central monitoring platform or on-site sensors through the control system. If an abnormal situation occurs (such as equipment failure, intrusion into a dangerous area, environmental changes, etc.), the control system will trigger an alarm according to the set emergency plan. The alarm will distinguish different types of warnings through high or low frequency sound, flashing or constant red, yellow, green light. For example, red flashing indicates a high-risk warning, yellow flashing indicates a matter of attention, and green indicates safe passage. This system can provide clear and direct warning information to on-site personnel visually and aurally, enhancing reaction speed and avoiding safety accidents.

[0103] Access control barrier is used to control the access channel of the construction site, and automatically adjusts the passage state according to the received switch instruction. By connecting with the personnel identity recognition system or automatic management system of the construction site, the access control barrier can control the opening or closing of the access channel according to different situations. The barrier can realize various identity verification methods such as card swiping, code scanning, and facial recognition to ensure that only authorized personnel can enter certain areas. For example, in areas where high-risk operations are required, only personnel with corresponding qualification certificates can pass through the barrier to enter. This not only improves the intelligent level of safety management, but also effectively monitors personnel access and prevents unauthorized personnel from entering dangerous areas.

[0104] The intelligent handheld terminal is a device that provides real-time instructions and information for field workers. This device can connect to the network to receive and display dispatch instructions, construction drawings, and safety warning information issued from the cloud. When there are new tasks or adjustments on the construction site, the cloud instructions will be pushed to each handheld terminal through the network, and workers can view the construction progress, drawing changes, and work requirements in real time. In addition, the intelligent terminal also supports the uploading of feedback information from field personnel to the edge computing node. This process can be done through voice recognition, taking pictures, or manual input, ensuring that on-site problems and abnormal conditions can be fed back to the central management system in a timely manner, promoting rapid response and processing. The edge computing node is responsible for preliminary processing and analysis of feedback data, improving reaction speed and decision-making efficiency, and avoiding data transmission delays affecting on-site operations.

[0105] The execution terminal in the BIM-based smart construction site management and control system provides efficient, safe, and intelligent management methods, not only improving the work efficiency of the construction site, but also strengthening the safety protection of personnel and equipment, and is an important technical facility indispensable for modern building construction management.

[0106] The present application encompasses any substitutions, modifications, equivalent methods and solutions made to the essence and scope of the present application. In order for the public to have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can be fully understood without these details by those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0107] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A BIM-based smart construction site management and control system, characterized in that, include: A cloud service platform, at least one edge computing node, and several on-site sensing and execution terminals; The on-site sensing and execution terminal is communicatively connected to the edge computing node, and the edge computing node is communicatively connected to the cloud service platform; The cloud service platform includes a BIM model integration and lightweight processing module, a multi-source data fusion and digital twin construction module, an artificial intelligence analysis and decision-making module, and a visualization interaction and command issuance module; the output end of the BIM model integration and lightweight processing module is connected to the input end of the multi-source data fusion and digital twin construction module. The output of the multi-source data fusion and digital twin construction module is connected to the input of the artificial intelligence analysis and decision-making module; the output of the artificial intelligence analysis and decision-making module is connected to the input of the visualization interaction and command issuance module; and the output of the visualization interaction and command issuance module is connected to the input of the edge computing node. The edge computing node includes a field data acquisition and preprocessing module, a local decision-making and real-time control module, and a communication management module. The output of the field data acquisition and preprocessing module is connected to the input of the local decision-making and real-time control module; The outputs of the field data acquisition and preprocessing module and the local decision-making and real-time control module are both connected to the input of the communication management module; the output of the communication management module is connected to the input of the cloud service platform. The on-site sensing and execution terminal includes a sensing terminal for collecting environmental status data and production factor status data, and an execution terminal for executing control commands.

2. The BIM-based smart construction site management system according to claim 1, characterized in that, The BIM model integration and lightweight processing module includes a model parsing unit and a lightweight generation unit; The model parsing unit is used to parse the file format of the original BIM model, extract the geometric information, attribute information and hierarchical relationship of the components, and identify the control-sensitive attributes of the components in the construction stage based on a predefined construction management knowledge base. The control-sensitive attributes include the construction process logic and safety management level to which the component belongs. The lightweight generation unit is data-connected to the model parsing unit and is used to dynamically determine the degree of model simplification based on terminal performance requirements and application scenarios. It calculates the importance score of components at the construction site, which is determined based on the component's control-sensitive attributes, its spatial location in the model, and its connection relationship with other components. Based on the determined importance score and degree of simplification, a mesh simplification algorithm is used to process the component's geometric information, removing internal invisible components and detailed features while retaining the component's unique identifier and key attributes, thereby generating a multi-level detail lightweight BIM model suitable for different terminals.

3. The BIM-based smart construction site management system according to claim 1, characterized in that, The multi-source data fusion and digital twin construction module includes a data fusion unit and a digital twin update unit; The data fusion unit is used to establish a unified spatiotemporal reference. It receives the actual measured coordinates of at least three reference points pre-set at the construction site and their coordinates in the BIM model, and uses a coordinate transformation algorithm to calculate the transformation parameters from the site coordinate system to the BIM model coordinate system. For all received sensor data, the coordinates are unified according to their equipment installation location and the transformation parameters. At the same time, the timestamps of all data sources are synchronized through the network time protocol to achieve spatiotemporal alignment of multi-source data. The twin update unit is connected to the data fusion unit and is used to update the spatiotemporally aligned data to the corresponding lightweight BIM model components according to the mapping relationship between the device ID and the components in the BIM model. When a data flow interruption or anomaly is detected, the data value of the missing period is calculated by linear interpolation or state estimation algorithm based on the historical data change pattern of the data source and other related sensor data, so as to ensure the continuity and integrity of the digital twin update.

4. The BIM-based smart construction site management system according to claim 1, characterized in that, The artificial intelligence analysis and decision-making module includes a progress prediction unit, a safety risk identification unit, and a scheduling optimization unit; The progress prediction unit is used to construct a dual-stream neural network model, where one stream inputs planned progress sequence data and the other stream inputs real-time acquired image recognition results, material arrival time sequence data, and personnel work hour data; it calculates the correlation weight between each feature in the real-time data stream and the planned progress feature through an attention mechanism, integrates the dual-stream features to predict the progress completion status at future time points, and identifies the delayed process that has the greatest impact on the critical path. The security risk identification unit is used to identify specific security risks using a multimodal fusion identification model, while simultaneously processing location data streams and video data streams. First, the precise location of the target personnel or equipment in the digital twin and the type of area it belongs to are determined through location data. At the same time, the video images of the corresponding area are analyzed through a convolutional neural network to extract security features. Finally, based on the decision-level fusion algorithm, the spatial location information and visual feature information are combined to output the security risk identification results and confidence level; The scheduling optimization unit is used to establish an optimization model based on the progress prediction results and safety risk identification results, with the goal of minimizing the construction period, minimizing the cost, or maximizing the resource utilization rate, and to use heuristic algorithms to solve for the optimal personnel, machinery, and material scheduling scheme.

5. A BIM-based smart construction site management system according to claim 1, characterized in that, The on-site data acquisition and preprocessing module includes a sensor network access unit, an image acquisition unit, and a data preprocessing unit; The sensor network access unit is used to support multiple industrial communication protocols and configure data acquisition frequencies and transmission formats that match different types of sensors. The image acquisition unit is used to connect the high-definition camera and the drone equipment, and to control its shooting angle, focal length and shooting interval. The data preprocessing unit is connected to the sensor network access unit and the image acquisition unit. It is used to clean the acquired raw data, remove outliers that are significantly beyond the physical range, and use a moving average filtering algorithm to smooth data fluctuations. For video stream data, it uses background subtraction and a lightweight target detection algorithm to identify moving targets and their basic features, and generate structured event description information. The processed sensor data and event description information are packaged together to form a preprocessed data package.

6. A BIM-based smart construction site management system according to claim 1, characterized in that, The local decision-making and real-time control module includes a rule engine unit and a real-time control unit; The rule engine unit is used to store and manage the security rule base distributed by the cloud. Each rule includes triggering conditions, execution actions, rule priority, and validity period. By continuously matching preprocessed real-time data with rule triggering conditions, the corresponding rule is triggered once a match is successful. The real-time control unit is connected to the rule engine unit and is used to receive the execution instructions of the triggered rules, generate corresponding control signals according to the instructions, and directly drive the on-site audible and visual alarms, access controllers or mechanical braking devices to perform corresponding actions, achieving millisecond-level response.

7. A BIM-based smart construction site management system according to claim 6, characterized in that, The rule engine unit also has a rule learning mechanism. When it detects that locally triggered rules and decision instructions issued by the cloud conflict multiple times within a certain period of time, it automatically records the context data of the conflict event, including environmental status, equipment status and operator information, and uploads this data to the cloud service platform as training data for rule optimization.

8. A BIM-based smart construction site management system according to claim 1, characterized in that, The communication management module includes a data uploading unit and an instruction receiving and distribution unit; The data upload unit is used to compress preprocessed data packets and local event data to be uploaded, using different compression algorithms according to the data type and urgency. For data containing personal privacy information, based on the privacy protection level assessment results, differential privacy technology is used to add random noise of appropriate intensity for desensitization processing; The instruction receiving and distribution unit is used to receive instruction data packets from the cloud service platform, parse the target device address and execution content of the instruction, verify the integrity and security of the instruction, and then distribute it to the local decision-making and real-time control module or the designated field execution terminal.

9. A BIM-based smart construction site management system according to claim 1, characterized in that, The sensing terminal includes personnel positioning tags, machinery operation status sensors, environmental monitoring sensors, and image acquisition devices; The personnel positioning tag uses UWB technology, and the two-dimensional or three-dimensional coordinates of the tag are calculated by measuring the arrival time of the tag signal through positioning base stations deployed at the construction site. The mechanical operating status sensors include tilt sensors, weight sensors, amplitude sensors and wind speed sensors, which are used to collect real-time operating parameters of mechanical equipment; The environmental monitoring sensors include a particulate matter concentration sensor, a noise sensor, and a temperature and humidity sensor, which are used to collect environmental data at the construction site. The image acquisition equipment includes a fixed high-definition dome camera and a mobile drone, used to acquire live video stream data.

10. A BIM-based smart construction site management system according to claim 1, characterized in that, The execution terminal includes an audible and visual alarm, an access control gate, and a smart handheld terminal; The sound and light alarm triggers different frequencies of sound and different colors of light warning according to the received control signal; The access control gate controls the passage status of the channel according to the received opening and closing instructions; The intelligent handheld terminal is used to receive and display scheduling instructions, construction drawings and safety warning information issued from the cloud, and to upload feedback information from on-site personnel to the edge computing node.

Images