BIM building intelligent integrated management system based on Internet of Things control

Abstract

The invention discloses a BIM building intelligent integrated management system based on Internet of Things control, and relates to the technical field of networking technology and building intelligent management, and the system comprises a sensing layer, a transmission layer, a platform layer and an application layer. The sensing layer comprises a multi-source heterogeneous sensor cluster and is divided into an environment monitoring unit, an equipment state acquisition unit, a personnel positioning unit and a structure health monitoring unit, and each unit is provided with a corresponding sensor; the transmission layer adopts a LoRaWAN and 5G fusion architecture and comprises an edge computing gateway, a TSN switch and a data security encryption unit, and the gateway can perform frequency domain analysis and the like; the platform layer is based on micro-service and is provided with a multi-source data fusion engine and a BIM model dynamic updating module; the application layer provides an equipment fault prediction module, an energy optimization module and a safety risk early warning module, and all the modules adopt corresponding algorithms. According to the invention, a hierarchical collaborative architecture is constructed, a multi-source sensor, a heterogeneous network and the like are integrated, deep integration of multi-dimensional management of a building is realized, and the problems of subsystem independence and data obstruction of a traditional system are solved.

Description

Technical Field

[0001] This invention relates to the fields of networking technology and intelligent building management technology, specifically to a BIM-based intelligent building integrated management system controlled by the Internet of Things. Background Technology

[0002] With the intelligent development of the construction industry, building management systems are gradually evolving towards integration and digitalization, but existing technologies still have many shortcomings:

[0003] Existing building management systems often operate with single functional modules running independently. Subsystems such as environmental monitoring, equipment maintenance, personnel management, and structural monitoring function in isolation, lacking a unified architecture. This results in incompatible data formats, poor information exchange, and difficulty in achieving collaborative management of the overall building's operational status. For example, equipment status data and environmental parameters cannot be linked for analysis, and personnel location information lacks real-time correlation with safety control measures, limiting the comprehensiveness and timeliness of management decisions.

[0004] Traditional solutions have significant limitations in terms of safety management and structural protection. Personnel positioning often relies on single technologies, resulting in insufficient accuracy and susceptibility to environmental obstructions, making precise tracking difficult in complex building spaces. Safety early warning systems rely heavily on manual inspections or fixed sensors, leading to delayed responses to emergencies such as unauthorized entry into restricted areas or unexpected personnel situations, and weak collaborative handling capabilities. Building structural health monitoring primarily employs offline analysis methods, failing to integrate real-time strain data with Building Information Modeling (BIM) for dynamic assessment, thus hindering the rapid development of targeted structural maintenance plans.

[0005] Technical bottlenecks also exist in the fields of energy management and data security. The regulation of energy systems such as HVAC is mostly based on preset parameters, failing to incorporate dynamic factors such as building usage and the external environment for intelligent optimization, leading to widespread energy waste. Simultaneously, the massive amounts of data generated during building management face challenges in transmission and storage; traditional encryption technologies suffer from issues such as easy key leakage and difficulty in tracing data tampering, failing to meet the high data security requirements of intelligent systems.

[0006] Therefore, there is an urgent need for an intelligent building integrated management solution that can achieve multi-dimensional data fusion, collaborative control of all aspects, and a balance between safety and energy efficiency, in order to solve the problems of low integration, slow response, low energy efficiency and insufficient data security in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a BIM-based intelligent building integrated management system based on Internet of Things control. This invention constructs a hierarchical collaborative architecture, integrates multi-source sensors, heterogeneous networks, etc., to achieve deep integration of multi-dimensional building management, solves the problems of independent subsystems and data disconnection in traditional systems, improves the overall management and intelligence, is more timely and reliable in security control, optimizes energy efficiency, and has full-link data security protection.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On one hand, a BIM-based intelligent building integrated management system controlled by the Internet of Things, the system comprising:

[0009] The perception layer, a cluster of multi-source heterogeneous sensors deployed within the building, includes:

[0010] The environmental monitoring unit includes temperature and humidity sensors, a CO2 sensor, and a PM2.5 sensor.

[0011] The equipment status acquisition unit includes a vibration sensor, a current transformer, and an infrared thermal imager.

[0012] Personnel positioning unit, including UWB positioning tag and BLE beacon;

[0013] Structural health monitoring unit, including strain gauges and tilt sensors;

[0014] The transport layer adopts a heterogeneous network architecture that integrates LoRaWAN and 5G, including:

[0015] An edge computing gateway has a built-in dynamic adaptive protocol conversion module and a real-time data analysis unit. The real-time data analysis unit is configured to perform frequency domain analysis and degradation index calculation of device status data.

[0016] Time-Sensitive Networking (TSN) switches enable millisecond-level data priority scheduling;

[0017] The data security encryption unit integrates a lightweight quantum key distribution module;

[0018] The platform layer, based on a microservice architecture, is an intelligent processing platform, including:

[0019] The multi-source data fusion engine performs the following operations: spatiotemporal alignment of the data stream from the perception layer and noise reduction using Kalman filtering; extracts device operation feature vectors and constructs a parameter anomaly matrix.

[0020] The BIM model dynamic update module maps real-time data to spatial topology nodes of the building information model.

[0021] The application layer provides the following service modules:

[0022] The equipment failure prediction module uses an LSTM neural network to analyze the parameter anomaly matrix;

[0023] The energy optimization module generates dynamic scheduling strategies for HVAC systems based on reinforcement learning algorithms.

[0024] The safety risk early warning module uses computer vision to identify unauthorized personnel entering construction restricted areas.

[0025] Furthermore, the personnel positioning unit includes:

[0026] A piezoelectric fall detection sensor embedded in a safety helmet collects attitude angular acceleration in real time;

[0027] The UWB / BLE dual-mode positioning tag achieves a three-dimensional spatial positioning accuracy of ±0.3m through the TDOA algorithm;

[0028] The edge computing gateway has a built-in inertial navigation compensation algorithm that reconstructs the trajectory using IMU data in areas with signal obstruction.

[0029] Furthermore, the real-time data analysis unit of the edge computing gateway performs:

[0030] Perform a sliding window Fourier transform on the device status data stream to extract fundamental frequency harmonic features;

[0031] When the harmonic distortion rate exceeds the threshold, the equipment degradation index calculation is triggered: Where w i The feature weights are w. i The influence weights of each equipment state characteristic (such as fundamental frequency, harmonic distortion rate, etc.) on the degradation judgment are obtained by training using the analytic hierarchy process (AHP) combined with historical equipment fault data, and satisfy the following conditions: S k,i S is the current eigenvalue. 0,i n1 is the initial calibration value, and n1 is the total number of equipment condition characteristics (such as the number of characteristics such as fundamental frequency and harmonic distortion rate) involved in the deterioration index calculation.

[0032] Deterioration Index (DI) k When the value is greater than 0.8, a device shutdown and maintenance command is pushed to the platform layer.

[0033] Furthermore, the multi-source data fusion engine includes:

[0034] The spatiotemporal alignment unit adopts an NTP / PTP hybrid clock synchronization protocol to control the timestamp deviation of RFID location data, equipment vibration data, and video stream within ±10ms;

[0035] The feature relation matrix construction unit performs the following steps: (a) generating the parameter anomaly matrix M. anomalyM anomaly This is an m×n matrix, where rows m correspond to the monitoring parameter type (such as equipment vibration value, temperature and humidity, etc.), and columns n correspond to time series sampling points; matrix elements M anomaly (i,j) is the anomaly index of the i-th type of parameter at time j, with a value range of [0,1] (0 represents normal, 1 represents severe anomaly), which is obtained by calculating the deviation between the parameter value after Kalman filtering and the dynamic threshold; (b) the envelope of the interface transmission signal is extracted by Hilbert transform, and the fluctuation amplitude coefficient is calculated. (σ is the standard deviation, μ is the mean); (c) Establish M anomaly With C f The mapping matrix R map ;

[0036] The dynamic threshold update unit automatically adjusts the anomaly detection boundary based on historical data clustering analysis.

[0037] Furthermore, the security risk warning module performs the following operations:

[0038] A three-dimensional electronic fence for construction restricted areas was constructed using lidar point cloud data.

[0039] When the UWB positioning tag enters the electronic fence, facial recognition is activated to verify permissions.

[0040] Triggering linkage control for unauthorized personnel:

[0041] Send a directional acoustic alarm to the speaker closest to the target location;

[0042] Control the drone to track the target and transmit video streams in real time;

[0043] Based on the YOLOv7 model, the helmet wearing status is identified, and the operation permissions of high-risk equipment are frozen when the helmet is not worn.

[0044] Furthermore, the energy optimization module performs:

[0045] Collect time-of-use electricity price data, weather forecast data, and building occupancy distribution;

[0046] Construct a multi-objective optimization function for the HVAC system: min(α·C) energy +β·C carbon +γ·∣T in -T set |), where C energy For energy consumption costs, C carbonFor carbon emission equivalents, α, β, and γ are the weighting coefficients for energy cost, carbon emissions, and indoor-outdoor temperature difference, respectively, satisfying α + β + γ = 1; where α is adjusted based on the time-of-use electricity price fluctuation coefficient, β is set with reference to regional carbon emission policy requirements, and γ is determined according to the comfort standards of building functions (such as office areas and residential areas). in T represents the real-time indoor temperature of a building. set Set the temperature for the HVAC system;

[0047] The NSGA-II algorithm is used to solve the Pareto optimal solution set, and the start-up and shutdown strategy of the chiller unit is dynamically adjusted.

[0048] Furthermore, the equipment fault prediction module:

[0049] Received parameter anomaly matrix M anomaly and fluctuation amplitude coefficient C f ;

[0050] Extract device degradation feature vectors using a temporal convolutional network;

[0051] The feature vectors are input into the dual-path LSTM model: Path 1: learns short-term (<24h) runtime abrupt change patterns; Path 2: captures long-term (>30 days) performance degradation trends;

[0052] The probability distribution function of the remaining lifespan of the output device.

[0053] Furthermore, when the structural health monitoring unit detects that the beam strain exceeds a threshold, it automatically executes:

[0054] Suspend the operation of construction equipment in the corresponding area;

[0055] Control the inspection robot to arrive at the site and collect high-definition images of cracks;

[0056] The strain data and BIM model were used to perform finite element analysis to calculate the structural safety factor.

[0057] If the safety factor is less than 1.5, send a 3D visualization command for the reinforcement plan to the management personnel.

[0058] Furthermore, the data security encryption unit performs:

[0059] Dynamic encryption key pairs are generated using a quantum key distribution module;

[0060] The data in the perception layer is encrypted end-to-end using the national cryptographic SM4 algorithm;

[0061] A blockchain-based evidence storage unit is set up at the platform layer to write abnormal device operation records into the Hyperledger Fabric distributed ledger.

[0062] On the other hand, a BIM-based intelligent building integrated management method based on Internet of Things (IoT) control includes:

[0063] Step 1: Mark the installation coordinates of IoT devices in the BIM model and generate a device topology diagram;

[0064] Step 2: Map physical devices to virtual models using a digital twin engine;

[0065] Step 3: When a new sensor is connected, execute the following automatically:

[0066] Protocol adaptation: Invoke the protocol library to match the communication protocol;

[0067] Data lineage tracing: marking the transmission path from the data source to the final application;

[0068] Step 4: Train a federated learning model based on historical fault data and iteratively optimize the warning threshold.

[0069] Compared with existing technologies, this IoT-based BIM building intelligent integrated management system has the following advantages:

[0070] I. This invention constructs a layered collaborative architecture of perception layer, transmission layer, platform layer, and application layer, integrating multi-source heterogeneous sensor clusters, heterogeneous network transmission, multi-source data fusion, and intelligent application modules. This achieves deep integration of multi-dimensional management of building environment, equipment, personnel, and structure. Compared with the shortcomings of traditional building management systems where subsystems operate independently and data is difficult to communicate, this system, with the help of the real-time analysis capabilities of edge computing gateways, spatiotemporal alignment and BIM dynamic mapping of the platform layer, and closed-loop control functions of the application layer, forms a complete link from data acquisition to intelligent decision-making, significantly improving the overall integrity and intelligence level of building management.

[0071] Second, this invention utilizes a personnel positioning unit to achieve precise personnel tracking and safety status monitoring through UWB / BLE dual-mode positioning and fall detection sensors. The safety risk early warning module combines a three-dimensional electronic fence, facial recognition, and multi-device linkage to construct an active defense mechanism. When the beam strain exceeds the limit, the structural health monitoring unit achieves dynamic assessment and handling of structural safety through equipment linkage, inspection robot detection, and BIM finite element analysis. This solves the problems of delayed response and incomplete coverage in the traditional manual inspection mode, and significantly improves the timeliness and reliability of building safety management.

[0072] Third, this invention effectively improves energy utilization efficiency by dynamically adjusting the HVAC system strategy based on the actual operating status of the building through the multi-objective optimization function and intelligent algorithm of the energy optimization module. At the same time, the data security encryption unit adopts quantum key distribution, national cryptographic algorithm encryption and blockchain evidence storage to build a full-link data security guarantee mechanism, which solves the problems of extensive energy regulation and prominent data transmission and storage security risks in traditional building management systems, and provides efficient and secure technical support for intelligent building management.

[0073] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0074] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0075] Figure 1 This is a diagram illustrating the overall architecture of a BIM-based intelligent building integrated management system controlled by the Internet of Things.

[0076] Figure 2 A flowchart for personnel positioning and safety early warning in a BIM-based intelligent building integrated management system controlled by the Internet of Things;

[0077] Figure 3 This is a flowchart of the equipment fault prediction data processing for a BIM-based intelligent building integrated management system controlled by the Internet of Things. Detailed Implementation

[0078] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0079] (I) Sensor Cluster Deployment in the Perception Layer

[0080] The perception layer achieves comprehensive monitoring of the building environment, equipment, personnel, and structure through a cluster of multi-source heterogeneous sensors, with the specific deployment as follows:

[0081] Environmental monitoring unit:

[0082] Sensors are deployed in each functional area of ​​the building as follows:

[0083] Temperature and humidity sensor: SHT30 model is selected, with a measurement range of -40℃~125℃ and 0~100%RH, and an accuracy of ±0.3℃ / ±2%RH. One unit is installed for every 50㎡, with a sampling cycle of 1 minute / time. It is mainly deployed in densely populated areas such as offices and conference rooms.

[0084] CO2 sensor: MH-Z19B model, measurement range 0~5000ppm, accuracy ±50ppm, installed in underground garages or poorly ventilated areas, sampling cycle 30 seconds / time;

[0085] PM2.5 sensor: PMS7003 model selected, measurement range 0~1000μg / m³ 3 Resolution 1μg / m 3 The monitoring equipment is deployed at the fresh air inlet and outdoor monitoring points, with a sampling cycle of 1 minute per time.

[0086] Equipment status acquisition unit:

[0087] Condition monitoring of critical equipment within the building:

[0088] Vibration sensor: A piezoelectric accelerometer (model YD-121) is used, with a measurement range of 0-50g and a frequency response of 1-10kHz. It is attached to the bearing housing of rotating equipment such as chillers and water pumps, with a sampling period of 10ms / time.

[0089] Current transformer: LMZJ1-0.5 type, transformation ratio 200 / 5A, accuracy class 0.5, installed at the outgoing end of the distribution cabinet, used to collect the operating current of the equipment, with a sampling period of 50ms / time;

[0090] Infrared thermal imager: HT-18 model, resolution 256×192, temperature measurement range -20℃~300℃, accuracy ±2%, for non-contact temperature measurement of equipment such as transformers and switch cabinets, with a sampling cycle of 1 minute / time.

[0091] Personnel positioning unit:

[0092] Real-time personnel location tracking and safety monitoring:

[0093] Safety helmets with integrated UWB / BLE dual-mode tags are provided for construction and maintenance personnel. The tags use DW1000 chip (UWB) and CC2541 chip (BLE), with a communication distance of 0-80m, a positioning sampling rate of 10Hz, and a three-dimensional positioning accuracy of ±0.3m achieved through the TDOA algorithm.

[0094] A piezoelectric fall detection sensor (model KX122) is embedded inside the safety helmet. It has a measurement range of ±2g and an output sensitivity of 6.1mg / LSB. It collects the wearer's angular acceleration in real time. When a fall is detected (acceleration change >1.5g and angle shift >45°), an alarm is triggered immediately.

[0095] Structural health monitoring unit:

[0096] Monitoring of key structural components of the building:

[0097] Strain gauges: BX120-5AA model, with a sensitivity coefficient of 2.10±1% and a resistance of 120Ω±0.1%, are selected and attached to the mid-span and supports of beams and columns. Four measuring points are set up for each beam, and the sampling cycle is 10 minutes / time.

[0098] Tilt sensor: SCA103T model is selected, with a measurement range of ±30° and an accuracy of ±0.01°. It is installed on the roof truss and at the corner of the floor, with a sampling cycle of 5 minutes / time.

[0099] (II) Heterogeneous Network Implementation at the Transport Layer

[0100] The transport layer adopts an architecture that integrates LoRaWAN and 5G, and the specific implementation is as follows:

[0101] Edge computing gateway:

[0102] An industrial-grade gateway (model EG-8000) was selected, configured with an ARM Cortex-A7 processor (1.2GHz), 1GB DDR3 memory, and supports LoRaWAN / 5G / Ethernet interfaces to achieve the following functions:

[0103] Dynamic adaptive protocol conversion: Built-in protocol library (including 12 protocols such as Modbus, MQTT, OPCUA, etc.) automatically matches the communication protocol through the device ID, such as using the Modbus-RTU protocol for vibration sensors and the MQTT protocol for positioning tags;

[0104] Real-time data analysis:

[0105] 1. Perform a sliding window Fourier transform on the equipment status data (such as vibration signals) (window size 512 points, overlap rate 50%) to extract the fundamental frequency and 2nd-5th harmonic features;

[0106] 2. When the harmonic distortion rate is >5%, calculate the degradation index: Where w i For characteristic weights (e.g., fundamental frequency w1 = 0.4, second harmonic w1 = 0.4), S k,i S is the current eigenvalue. k,0 This is the initial calibration value; when DI kWhen the value is greater than 0.8, a shutdown maintenance command is sent to the platform layer;

[0107] Inertial navigation compensation: In areas with signal obstruction, such as elevator shafts, the trajectory is reconstructed using IMU (Inertial Measurement Unit) data (sampling rate 100Hz) to compensate for positioning errors to within ±0.5m.

[0108] Time-Sensitive Networking (TSN) Switch

[0109] Deploy a TSN switch (model TS-6008) with 8 Gigabit Ethernet ports, supporting the IEEE 802.1AS synchronization protocol. Data is prioritized for scheduling: critical data such as equipment fault warnings and structural strain are given the highest priority, with a transmission latency ≤10ms; environmental monitoring and personnel positioning data are given secondary priority, with a transmission latency ≤100ms. A data security encryption unit is also included.

[0110] Achieving end-to-end data security protection: The quantum key distribution module (model QKD-200) generates a 256-bit dynamic key, which is updated every 30 minutes; the national cryptographic SM4 algorithm is used to encrypt the data in the perception layer (128-bit block length, 32 rounds of encryption) to achieve end-to-end encryption; the platform layer deploys Hyperledger Fabric blockchain nodes (3 consensus nodes) to write abnormal device operation records (such as unauthorized start / stop, parameter modification) into the distributed ledger, with a block generation interval of 60 seconds and each block containing 50 records.

[0111] (III) Intelligent Processing Implementation at the Platform Layer The platform layer is built on a microservice architecture (using the Spring Boot framework) and deployed on two physical servers (8 cores, 16GB memory). Specific module implementations are as follows: Multi-source data fusion engine spatiotemporal alignment: NTP / PTP hybrid synchronization is used (NTP for non-real-time data, accuracy ±10ms; PTP for real-time data, accuracy ±1ms), calibrating the timestamps of RFID, vibration, and video streams to ensure that the time deviation of the same event is ≤10ms; Feature relationship matrix construction:

[0112] 1. Generate the parameter anomaly matrix M anomaly (m×n matrix): Row m corresponds to 10 types of parameters (such as vibration, current, temperature and humidity), column n corresponds to time sampling points (3600 points per hour), and element M anomaly (i,j) is the anomaly index (range [0,1]), which is calculated by the deviation between the value after Kalman filtering and the dynamic threshold. For example, if a certain temperature parameter is 30℃ after filtering, the upper limit of the dynamic threshold is 26℃, and the severe threshold is 32℃, then the anomaly index = (30-26) / (32-26) = 0.67.

[0113] 2. Extract the signal envelope using Hilbert transform and calculate the fluctuation amplitude coefficient. (σ is the standard deviation of the signal over 5 minutes, and μ is the mean);

[0114] 3. Establish M anomaly With C f The mapping matrix R map The correlation was calculated using the Pearson coefficient.

[0115] Dynamic threshold update: Based on K-means clustering (k=3) analysis of historical data over the past 3 months, the threshold is adjusted monthly (e.g., the summer temperature threshold is automatically increased by 2℃).

[0116] BIM model dynamic update module

[0117] A BIM model (including building structure, equipment, and sensor coordinates) is created using Revit. Real-time data (such as equipment current and ambient humidity) is mapped to the model topology nodes via an API interface, with an update frequency of 1 minute. The status is displayed using color coding (green: normal, yellow: slight abnormality, red: severe abnormality).

[0118] (iv) Implementation of application layer service modules

[0119] Equipment Failure Prediction Module:

[0120] Receive M anomaly and C f A 64-dimensional degenerate feature vector is extracted using a temporal convolutional network (3 convolutional layers, kernel sizes 3 / 5 / 7);

[0121] Input a dual-path LSTM model: Path 1 (2-layer LSTM) learns short-term mutations <24h, Path 2 (2-layer LSTM) captures long-term decays >30 days, and outputs the probability distribution of remaining lifespan (e.g., water pump remaining lifespan 18±2 days, probability 65%).

[0122] Energy optimization module:

[0123] Data collection: Time-of-use electricity price (peak hour 1.2 yuan / kWh, normal hour 0.7 yuan / kWh, off-peak hour 0.3 yuan / kWh), 24-hour weather forecast (obtained from meteorological API), and population distribution (statistical data collected via UWB location tracking);

[0124] Multi-objective optimization function: min(α·C) energy +β·C carbon +γ·∣T in -T set |) where α=0.4 (adjusted to 0.6 during peak hours), β=0.3, γ=0.3 (γ=0.4 in office area), C energy For energy consumption costs, C carbon T is the carbon emission equivalent (kgCO2).in For real-time indoor temperature, T set Set the temperature (26℃ for the office area);

[0125] The NSGA-II algorithm (population 100, iterations 50) is used to solve the problem and dynamically adjust the start and stop of the chiller units (e.g., 3 units are turned on during off-peak hours and 1 unit is turned on during peak hours).

[0126] Security risk warning module

[0127] 3D electronic fence: Constructing construction restricted areas (e.g., 5m around a foundation pit) using point cloud data from LiDAR (RS-LIDAR-16);

[0128] Permission verification: When the UWB tag enters the fence, it triggers face recognition by the camera (4 megapixels) (ArcFace algorithm, accuracy 99.7%).

[0129] Linkage control: When unauthorized personnel break in, a 1kHz alarm is sent to the nearest speaker (110dB), and a drone (with a flight time of 30 minutes) is controlled to track and transmit video back. The safety helmet is identified by the YOLOv7 model (training set contains 8000 images), and the tower crane operation permission is frozen if the helmet is not worn.

[0130] Structural health linkage control

[0131] When the beam strain exceeds the threshold (2000με):

[0132] Suspend the operation of equipment (such as concrete pump trucks) in the area;

[0133] An inspection robot (equipped with a 20-megapixel camera) collects images of cracks along a track;

[0134] Import the BIM model into ANSYS for finite element analysis and calculate the safety factor. If it is less than 1.5, push the AR reinforcement command (carbon fiber cloth bonding position / thickness).

[0135] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A BIM-based intelligent building integrated management system controlled by the Internet of Things, characterized in that, The system includes: The perception layer, a cluster of multi-source heterogeneous sensors deployed within the building, includes: The environmental monitoring unit includes temperature and humidity sensors, a CO2 sensor, and a PM2.5 sensor. The equipment status acquisition unit includes a vibration sensor, a current transformer, and an infrared thermal imager. Personnel positioning unit, including UWB positioning tag and BLE beacon; Structural health monitoring unit, including strain gauges and tilt sensors; The transport layer adopts a heterogeneous network architecture that integrates LoRaWAN and 5G, including: An edge computing gateway has a built-in dynamic adaptive protocol conversion module and a real-time data analysis unit. The real-time data analysis unit is configured to perform frequency domain analysis and degradation index calculation of device status data. Time-Sensitive Networking (TSN) switches enable millisecond-level data priority scheduling; The data security encryption unit integrates a lightweight quantum key distribution module; The platform layer, based on a microservice architecture, is an intelligent processing platform, including: The multi-source data fusion engine performs the following operations: spatiotemporal alignment of the data stream from the perception layer and noise reduction using Kalman filtering; extracting device operation feature vectors and constructing a parameter anomaly matrix. The BIM model dynamic update module maps real-time data to spatial topology nodes of the building information model. The application layer provides the following service modules: The equipment failure prediction module uses an LSTM neural network to analyze the parameter anomaly matrix; The energy optimization module generates dynamic scheduling strategies for HVAC systems based on reinforcement learning algorithms. The safety risk early warning module uses computer vision to identify unauthorized personnel entering construction restricted areas.

2. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The personnel positioning unit includes: A piezoelectric fall detection sensor embedded in a safety helmet collects attitude angular acceleration in real time; The UWB / BLE dual-mode positioning tag achieves a three-dimensional spatial positioning accuracy of ±0.3m through the TDOA algorithm; The edge computing gateway has a built-in inertial navigation compensation algorithm that reconstructs the trajectory using IMU data in areas with signal obstruction.

3. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The real-time data analysis unit of the edge computing gateway performs the following: Perform a sliding window Fourier transform on the device status data stream to extract fundamental frequency harmonic features; When the harmonic distortion rate exceeds the threshold, the equipment degradation index calculation is triggered: Where w i For feature weights, feature weight w i The influence weights of each equipment condition characteristic on the degradation judgment are obtained by training using the analytic hierarchy process (AHP) combined with historical equipment failure data, and satisfy the following conditions: S k,i S is the current eigenvalue. 0,i n1 is the initial calibration value, and n1 is the total number of equipment condition characteristics involved in the deterioration index calculation. Deterioration Index (DI) k When the value is greater than 0.8, a shutdown and maintenance command is pushed to the platform layer.

4. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The multi-source data fusion engine includes: The spatiotemporal alignment unit adopts an NTP / PTP hybrid clock synchronization protocol to control the timestamp deviation of RFID location data, equipment vibration data, and video stream within ±10ms; The feature relation matrix construction unit performs the following steps: (a) generating the parameter anomaly matrix M. anomaly M anomaly It is an m×n matrix, where row m corresponds to the monitoring parameter type and column n corresponds to the time series sampling points; matrix elements M anomaly (i,j) is the anomaly index of the i-th type of parameter at time j, with a value range of [0,1] (0 represents normal, 1 represents severe anomaly), which is obtained by calculating the deviation between the parameter value after Kalman filtering and the dynamic threshold; (b) the envelope of the interface transmission signal is extracted by Hilbert transform, and the fluctuation amplitude coefficient is calculated. (σ is the standard deviation, μ is the mean); (c) Establish M anomaly With C f The mapping matrix R map ; The dynamic threshold update unit automatically adjusts the anomaly detection boundary based on historical data clustering analysis.

5. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The security risk warning module performs the following operations: A three-dimensional electronic fence for construction restricted areas was constructed using lidar point cloud data. When the UWB positioning tag enters the electronic fence, facial recognition is activated to verify permissions. Triggering linkage control for unauthorized personnel: Send a directional acoustic alarm to the speaker closest to the target location; Control the drone to track the target and transmit video streams in real time; Based on the YOLOv7 model, the helmet wearing status is identified, and the operation permissions of high-risk equipment are frozen when the helmet is not worn.

6. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The energy optimization module performs the following: Collect time-of-use electricity price data, weather forecast data, and building occupancy distribution; Construct a multi-objective optimization function for the HVAC system: min(α·C) energy +β·C carbon +γ·∣T in -T set |), where C energy For energy consumption costs, C carbon For carbon emission equivalents, α, β, and γ are the weighting coefficients for energy cost, carbon emissions, and indoor-outdoor temperature difference, respectively, satisfying α + β + γ = 1; where α is adjusted based on the time-of-use electricity price fluctuation coefficient, β is set with reference to regional carbon emission policy requirements, and γ is determined according to the comfort standards of building functions. in T represents the real-time indoor temperature of a building. set Set the temperature for the HVAC system; The NSGA-II algorithm is used to solve the Pareto optimal solution set, and the start-up and shutdown strategy of the chiller unit is dynamically adjusted.

7. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The equipment fault prediction module: Received parameter anomaly matrix M anomaly and fluctuation amplitude coefficient C f ; Extract device degradation feature vectors using a temporal convolutional network; The feature vectors are input into the dual-path LSTM model: Path 1: learns short-term (<24h) runtime abrupt change patterns; Path 2: captures long-term (>30 days) performance degradation trends; The probability distribution function of the remaining lifespan of the output device.

8. The BIM-based intelligent building integrated management system according to claim 5, characterized in that, When the structural health monitoring unit detects that the beam strain exceeds the threshold, it automatically executes the following: Suspend the operation of construction equipment in the corresponding area; Control the inspection robot to arrive at the site and collect high-definition images of cracks; The strain data and BIM model were used to perform finite element analysis to calculate the structural safety factor. If the safety factor is less than 1.5, send a 3D visualization command for the reinforcement plan to the management personnel.

9. The BIM-based intelligent building integrated management system according to claim 1, characterized in that, The data security encryption unit performs the following: Dynamic encryption key pairs are generated using a quantum key distribution module; The data in the perception layer is encrypted end-to-end using the national cryptographic SM4 algorithm; A blockchain-based evidence storage unit is set up at the platform layer to write abnormal device operation records into the Hyperledger Fabric distributed ledger.

10. A BIM building intelligent integrated management method based on Internet of Things (IoT) control, applicable to the BIM building intelligent integrated management system based on IoT control as described in claims 1-9, characterized in that... The method includes: Step 1: Mark the installation coordinates of IoT devices in the BIM model and generate a device topology diagram; Step 2: Map physical devices to virtual models using a digital twin engine; Step 3: When a new sensor is connected, execute the following automatically: Protocol adaptation: Invoke the protocol library to match the communication protocol; Data lineage tracing: marking the transmission path from the data source to the final application; Step 4: Train a federated learning model based on historical fault data and iteratively optimize the warning threshold.

Images