Building construction carbon emission real-time monitoring and intelligent analysis system based on Internet of Things

Abstract

The invention provides a building construction carbon emission real-time monitoring and intelligent analysis system based on the Internet of Things, relates to the technical field of carbon emission monitoring and analysis, and solves the problems of functional deficiency and imperfection of a carbon emission monitoring system. In the system, a sensing layer is used for collecting multi-dimensional carbon emission related original data of a construction site in real time and carrying out edge side preprocessing; the transmission layer is used for transmitting data by adopting a self-adaptive multi-mode communication protocol and adapting to the network environment change of a construction site; the platform layer is used for receiving and integrating data, executing carbon emission accounting, performing multi-dimensional intelligent analysis, constructing and updating a digital twinborn simulation model and driving an intelligent analysis process; and the application layer is used for visually and dynamically presenting system information data, providing interactive access and early warning, and opening key carbon emission data and supervision support to a government supervision platform. The method is optimized and improved from multiple aspects, so that the system has relatively high application efficiency and meets the requirements of domestic and overseas carbon auditing.

Description

Technical Field

[0001] This invention relates to the field of carbon emission monitoring and analysis technology, specifically to an Internet of Things-based real-time monitoring and intelligent analysis system for carbon emissions from building construction. Background Technology

[0002] As global dual-carbon goals are being pursued more deeply, the construction industry, as a significant source of carbon emissions, faces a critical challenge in achieving its green transformation through precise control of carbon emissions during the construction phase. However, current technological systems in this field still have significant limitations, hindering the achievement of emission reduction targets.

[0003] Current technologies for data acquisition remain weak. Data collection largely relies on manual recording or single-type sensors, employing rudimentary methods with limited coverage. This results in fragmented data, making it difficult to comprehensively capture indirect emission sources such as transportation fuel consumption and dust emissions, leading to significant omissions in the calculation scope. Furthermore, data updates are severely delayed, failing to reflect instantaneous emission peaks during construction and hindering timely management decision-making. In addition, the lack of reliable and effective methods for locating and tracking mobile sources such as transport vehicles distorts the basis for calculating related emissions.

[0004] Current carbon emission accounting methods also have significant shortcomings, especially in terms of accuracy, which urgently needs improvement. Mainstream methods generally employ static models and fixed emission factors, failing to fully consider dynamic changes such as equipment aging and fluctuations in actual operating conditions, leading to a disconnect between calculated emission intensity and reality. Furthermore, existing accounting systems struggle to trace emission data back to specific construction procedures or component levels within building information models, failing to identify critical high-emission construction nodes, severely hindering the formulation and implementation of precise emission reduction measures.

[0005] The existing systems also have a relatively low level of intelligence, lacking foresight and proactive optimization capabilities. Data analysis mostly remains at the level of simple statistics and display of historical data, unable to effectively predict future carbon emission trends or deeply diagnose the root causes of high emissions. The systems are generally in a passive response mode, unable to generate proactive optimization strategy suggestions based on real-time and predictive data, such as dynamic adjustment plans for machinery scheduling and construction sequence arrangements, making it difficult to form a closed-loop management process of "monitoring-analysis-optimization-verification".

[0006] Low system integration and poor compatibility are another major limitation of current technologies. Key subsystems such as energy consumption monitoring, construction progress management, and vehicle dispatching often operate in isolation, with varying data formats, forming persistent data silos that hinder cross-system data fusion analysis and collaborative applications. Existing solutions typically lack flexible architecture, often custom-developed for specific static scenarios, making it difficult to adapt to complex and ever-changing construction site environments, such as parallel operations on multiple work fronts and dynamic adjustments to construction plans, thus limiting the technology's potential for large-scale application and widespread adoption.

[0007] It is evident that existing technologies have systemic shortcomings in terms of the comprehensiveness of data collection, the accuracy of calculation, intelligent decision-making, and system integration, and innovative solutions are urgently needed to achieve truly refined and dynamic management of carbon emissions during the construction phase. Summary of the Invention

[0008] The purpose of this invention is to address the functional deficiencies and limitations of existing carbon emission monitoring systems for building construction processes. Therefore, it proposes an Internet of Things (IoT)-based real-time monitoring and intelligent analysis system for carbon emissions from building construction. This system has been optimized and improved in multiple aspects, including data acquisition, processing and transmission, calculation and prediction, and monitoring presentation. This results in high application efficiency and compliance with domestic and international carbon audit requirements, thereby achieving transparent and efficient management of carbon emission intensity and equipment energy efficiency.

[0009] The present invention employs the following technical solutions to achieve its objective: A real-time monitoring and intelligent analysis system for carbon emissions from building construction based on the Internet of Things (IoT), the system comprising: The perception layer is used to collect multi-dimensional raw data related to carbon emissions at the construction site in real time, including environmental parameters, energy consumption parameters, equipment operating parameters and material flow information, and to perform edge-side preprocessing on the raw data. The transport layer is used to encrypt and transmit pre-processed perception layer data to the platform layer using an adaptive multimode communication protocol, and to adapt to changes in the network environment at the construction site. The platform layer is used to receive and integrate data from the transport layer, perform carbon emission accounting based on real-time data, use artificial intelligence technology to conduct multi-dimensional intelligent analysis of the accounting results and related data, identify emission characteristics, anomalies and optimization potential, and build and update digital twin simulation models of the construction site based on the integrated data to drive the intelligent analysis process. The application layer is used to visualize and dynamically present the carbon emission accounting results, intelligent analysis conclusions, and digital twin simulation status information from the platform layer; it provides interactive access and early warning through mobile applications; and it opens up key carbon emission data and regulatory support to government regulatory platforms through standardized interfaces.

[0010] Specifically, the perception layer is configured with a heterogeneous sensor network, including: Energy consumption sensor arrays are used to collect device-level energy consumption data; Environmental sensor arrays are used to monitor combustion emissions and dust data; Positioning sensor arrays are used to track material transport paths.

[0011] Preferably, the perception layer is also equipped with edge computing nodes and BIM system data interfaces; The heterogeneous sensor network also includes a material consumption monitoring unit, which is used to obtain the amount of various types of building materials entering the site and the actual consumption in real time through the coordinated operation of RFID tags and weighing sensors. The material consumption monitoring unit is connected to the data interface of the BIM system. The material consumption monitoring unit is also used to compare and analyze the deviation between the actual consumption and the planned consumption in the model of the BIM system. When the deviation value exceeds the preset threshold, an early warning signal is triggered. Edge computing nodes have built-in lightweight CNN models, which are used to identify operating conditions and filter outliers from raw data of multiple types of sensors from heterogeneous sensor networks.

[0012] Specifically, lightweight CNN models include: The input layer is used to receive noise spectrum diagrams and vibration time-series data from raw data of various types of sensors. Convolutional layers are used to extract device operating condition features using kernels of a preset size, and the convolution stride is preset. The pooling layer is used to control the dimensionality reduction ratio to a preset value using the max pooling method. The output layer is used to classify and output the operating status of the equipment.

[0013] Preferably, the multimode communication protocol used in the transport layer includes: In pre-designed areas with high concentrations of construction machinery, a self-organizing network is constructed using the LoRa protocol; For mobile devices, the NB-IoT protocol is used, and their data packets are compressed using the COAP protocol.

[0014] Preferably, the transport layer is equipped with a device authentication module, which is used to authenticate each sensor device in the heterogeneous sensor network using a two-way authentication mechanism based on the national cryptographic SM2 algorithm; each sensor device is equipped with a unique digital certificate.

[0015] Preferably, the platform layer is configured with a carbon accounting engine and an AI analysis module, wherein: The carbon accounting engine is used to build dynamic calculation models based on the ISO 14064 standard and output carbon emission intensity indicators in Scope 1-3. The AI ​​analysis module integrates an LSTM prediction model, a random forest root cause analysis model, and a genetic algorithm optimizer. It is used to predict carbon emission trends and generate emission reduction strategies.

[0016] Specifically, LSTM prediction models include: The input layer is used to receive carbon emission time-series data, meteorological data, and construction progress data for a preset historical period. The hidden layer has a preset number of LSTM units and corresponding activation functions. The hidden layer is used to receive various data inputs from the input layer, calculate them, and output them to the output layer. The output layer is used to generate carbon emission predictions for a preset future time period. The LSTM prediction model is also used to perform incremental learning algorithms within a preset period to update its own model weight parameters; The genetic algorithm optimizer is used to output the optimal scheduling scheme by executing the execution logic; the execution logic is as follows: Initialize the population: Generate a preset number of machine scheduling schemes, each of which contains the start and stop schedules for each type of machine; Fitness calculation: Using the minimization of total carbon emissions as the objective function, calculate the fitness value for each group of machine scheduling schemes; Selection operation: Use roulette wheel selection to retain individuals with the highest fitness ranking in the first preset percentage; Crossover mutation: Perform single-point crossover and random mutation with preset probabilities on the retained individuals; Iteration termination condition: After a preset number of iterations, the optimal solution is output as the optimal scheduling scheme when the rate of change of the optimal solution is less than a second preset percentage.

[0017] Preferably, the platform layer is also equipped with a digital twin emulator, including: BIM model parsing unit is used to extract the component material properties and construction sequence logical relationships of the model in the BIM system; The carbon emission mapping unit is used to spatially bind real-time collected and monitored data with model components in the BIM system to form a 4D carbon emission distribution model. The scheme comparison unit is used to simulate the impact of different construction sequences on total carbon emissions and calculate and quantify the differences between the schemes.

[0018] Preferably, the application layer is configured with a large-screen visualization terminal, a mobile app, and a government regulatory API interface, wherein: Large-screen visualization terminals are used for the dynamic and visual presentation of information and data. The mobile app includes a foreman terminal and a management terminal. The foreman terminal is used to view the mechanical energy efficiency ranking and display the carbon emission index corresponding to the unit workload. The management terminal is used to present a comparison view of carbon emissions from multiple projects, generate and present deviation reports from the construction contract. The government regulatory API interface is used to encapsulate system data in JSON-LD format for government devices to call for supervision. The encapsulated system data fields include basic project information, carbon emission data, and audit trail information. The government regulatory API interface adopts the OAuth 2.0 authorization protocol.

[0019] In summary, due to the adoption of this technical solution, the beneficial effects of this invention are as follows: This invention effectively overcomes several key deficiencies in existing building carbon emission management, achieving refined and intelligent control of carbon emissions during the construction phase. It possesses comprehensive and accurate data collection and processing capabilities, dynamic carbon emission accounting and traceability capabilities, closed-loop optimization decision-making capabilities, and highly compatible safety and compliance features.

[0020] In terms of data acquisition and processing, this invention, based on its collaboratively deployed heterogeneous sensor network, achieves comprehensive real-time monitoring of multiple sources of factors, including energy consumption of mechanical equipment, transportation activity trajectories, and environmental emissions. This improves the completeness and timeliness of data coverage, changing the fragmented and lagging state of traditional reliance on manual or single data sources. Simultaneously, by leveraging the localized preprocessing capabilities of edge computing nodes, the burden of transmitting raw data to the cloud is reduced, ensuring real-time and efficient data processing and transmission.

[0021] In terms of carbon emission accounting and traceability, this invention combines international standards with dynamic construction data from Building Information Modeling (BIM) to construct a dynamic carbon accounting engine. This engine can accurately trace carbon emissions to specific construction procedures and material consumption stages, enabling in-depth analysis from the overall project level to the process level. This improves the accuracy and guiding value of the accounting, thus providing a reliable basis for precise emission reduction.

[0022] In terms of intelligent decision support, the system deeply integrates multiple algorithm models to construct a closed loop of prediction, analysis, and optimization. The system can not only proactively predict carbon emission trends but also deeply diagnose the root causes of high emissions and actively generate optimization strategies with significant emission reduction effects, such as dynamic adjustment schemes for machinery scheduling and work process arrangements. The system has continuous evolution capabilities and can quickly adapt to the needs of different construction site scenarios.

[0023] In terms of visualization and management, this invention spatiotemporally binds real-time monitoring data with building information models, generating a heat map that intuitively reflects the spatiotemporal distribution characteristics of carbon emissions, supporting dynamic simulation and comparison of different construction schemes. Through a multi-terminal hierarchical visualization interface, transparent and refined management of carbon emission intensity and equipment operating efficiency is achieved.

[0024] In terms of system integration and security compliance, this invention applies secure authentication and encrypted transmission technologies to ensure data security and reliability throughout the entire process, and the legality and trustworthiness of equipment. Through standardized data interfaces and format design, it seamlessly connects with government regulatory requirements and automatically generates audit reports that comply with authoritative domestic and international standards, thereby achieving a high degree of system compatibility and compliance. Attached Figure Description

[0025] The present invention will be further described in detail with reference to the following figures, specifically including one figure as follows: Figure 1 This is a schematic diagram of the structural composition of the carbon emission real-time monitoring and intelligent analysis system of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0027] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0028] A real-time monitoring and intelligent analysis system for carbon emissions from building construction based on the Internet of Things (IoT). Figure 1 The overall structure of the system is shown and can be viewed simultaneously. The key hierarchical architecture is as follows: The perception layer is used to collect multi-dimensional raw data related to carbon emissions at the construction site in real time, including environmental parameters, energy consumption parameters, equipment operating parameters and material flow information, and to perform edge-side preprocessing on the raw data. The transport layer is used to encrypt and transmit pre-processed perception layer data to the platform layer using an adaptive multimode communication protocol, and to adapt to changes in the network environment at the construction site. The platform layer is used to receive and integrate data from the transport layer, perform carbon emission accounting based on real-time data, use artificial intelligence technology to conduct multi-dimensional intelligent analysis of the accounting results and related data, identify emission characteristics, anomalies and optimization potential, and build and update digital twin simulation models of the construction site based on the integrated data to drive the intelligent analysis process. The application layer is used to visualize and dynamically present the carbon emission accounting results, intelligent analysis conclusions, and digital twin simulation status information from the platform layer; it provides interactive access and early warning through mobile applications; and it opens up key carbon emission data and regulatory support to government regulatory platforms through standardized interfaces.

[0029] As a preferred embodiment, the sensing layer is configured with a heterogeneous sensor network, including: Energy consumption sensor groups are used to collect equipment-level energy consumption data; such sensor groups may include smart meters, gas flow meters, diesel flow meters, etc., installed on construction machinery.

[0030] Environmental sensor arrays are used to monitor combustion emissions and dust data; these sensor arrays may include CO2 concentration sensors, PM2.5 sensors, and noise sensors deployed in construction areas.

[0031] A positioning sensor array is used to track material transport paths. This type of sensor array may include a GPS positioning module and a UWB positioning tag mounted on the transport vehicle. The UWB positioning tag achieves high-precision real-time positioning through ultra-wideband technology and can provide centimeter-level positioning accuracy in complex environments. It is suitable for material tracking scenarios with high positioning accuracy requirements, such as those in this embodiment.

[0032] As a preferred embodiment, the perception layer is also equipped with edge computing nodes and a BIM system data interface. The heterogeneous sensor network also includes a material consumption monitoring unit, which is used to obtain the arrival quantity and actual consumption of various types of building materials such as steel bars and concrete in real time through the coordinated operation of RFID tags and weighing sensors.

[0033] In this embodiment, the material consumption monitoring unit is connected to the BIM system data interface. The material consumption monitoring unit is also used to compare and analyze the deviation between the actual consumption and the planned consumption in the model of the BIM system. When the deviation value exceeds the preset threshold, an early warning signal is triggered.

[0034] The material consumption monitoring unit, integrated with the BIM system data interface, enables dynamic comparative analysis of material usage at the construction site against the planned usage in the design model. This sensing technology, based on RFID tags and weighing sensors, not only improves the precision of material management but also promptly detects deviations in material usage during construction, such as abnormal consumption due to construction errors or waste, all of which can potentially increase carbon emissions. When the system detects a deviation between actual consumption and the planned usage in the BIM model exceeding a preset threshold, it automatically triggers an early warning signal, alerting relevant management personnel to take appropriate measures. This effectively controls carbon emission configuration and enhances the overall carbon emission management level and informatization of the project.

[0035] In this embodiment, the edge computing node incorporates a lightweight convolutional neural network model, namely a lightweight CNN model. This model is specifically designed for real-time condition identification and outlier filtering of raw data from various types of sensors in a heterogeneous sensor network. By deploying this model at the edge, system communication overhead can be effectively reduced, data processing efficiency improved, and the real-time performance and reliability of equipment operating status monitoring enhanced.

[0036] The lightweight CNN model includes the following core components: 1. Input Layer: This layer receives multimodal raw data from a heterogeneous sensor network, specifically including: Noise Spectrum Data: a two-dimensional noise spectrum image acquired by an audio sensor and generated using methods such as Short-Time Fourier Transform (STFT); Vibration Time Series Data: raw vibration signals acquired by an accelerometer, represented in one-dimensional time series form. The input layer performs uniform formatting on the above different types of data to form input tensors adapted to the subsequent network structure.

[0037] 2. Convolutional Layers: Used to extract key equipment operation features from the input data. This implementation uses multiple small 3×3 convolutional kernels to scan the input data layer by layer in a sliding window manner, capturing local spatial correlations and temporal variation patterns. To improve feature extraction efficiency and control model complexity, the stride of the convolutional operation is set to 2, thereby achieving initial dimensionality reduction while maintaining feature resolution.

[0038] 3. Pooling Layer: Used to employ max pooling to further compress the spatial dimension of the feature map, reduce redundant information, and enhance model robustness. In this implementation, the pooling window size is set to match the stride, so that the size of the feature map is reduced to 1 / 4 of its original size after each pooling operation, thereby effectively controlling the overall computational load of the model.

[0039] 4. Output Layer: This layer uses the Softmax classification function to classify the high-order feature vectors extracted by the model. Specifically, the model outputs three classification results for the current operating state of the device: normal operating state, idling state (not working but running), and fault state (abnormal vibration or abnormal noise). The classification results are output in the form of a probability distribution for use by the subsequent decision-making module.

[0040] This implementation achieves efficient preprocessing and intelligent analysis of multi-source heterogeneous sensor data by deploying a lightweight CNN model on edge computing nodes, which facilitates the use of subsequent system hierarchical architecture.

[0041] As a preferred embodiment, the multimode communication protocol used by the transport layer includes: In a pre-defined area with dense construction machinery, a self-organizing network is constructed using the LoRa protocol, with a communication frequency band of 470-510MHz and a transmission distance of ≥800 meters. LoRa, a low-power, long-range wireless communication technology widely used in the Internet of Things (IoT) field, is suitable for this scenario. In areas with dense construction machinery, the presence of numerous metal structures and high-noise environments makes communication signals susceptible to interference; therefore, using a self-organizing network enhances network robustness and coverage. The 470-510MHz frequency band has good penetration and propagation distance, making it suitable for complex terrain and densely built-up areas.

[0042] For mobile devices, the NB-IoT protocol is adopted, and its data packets are compressed using the COAP protocol, with a compression rate of ≥60%. The NB-IoT protocol enables mobile devices to achieve stable wide area network access, while the COAP protocol improves its communication efficiency and reduces energy consumption. Ensuring the compression rate means reducing the amount of data transmitted by optimizing the data format and reducing redundant fields, thereby saving communication costs and energy consumption.

[0043] Meanwhile, the transport layer is equipped with a device authentication module. This module uses a two-way authentication mechanism based on the Chinese national cryptographic algorithm SM2 to authenticate each sensor device in the heterogeneous sensor network. Each sensor device is equipped with a unique digital certificate. The SM2 algorithm is an elliptic curve public key cryptography standard issued by the State Cryptography Administration of China, which has high security and complies with domestic information security standards. Each sensor device has a unique digital certificate, equivalent to its "ID card," issued by a trusted certificate authority. The digital certificate contains the device's public key, identity information, and signature, and is the foundation for implementing SM2 authentication.

[0044] As a preferred embodiment, the platform layer is configured with a carbon accounting engine and an AI analysis module, wherein: The carbon accounting engine is used to build dynamic calculation models based on the ISO 14064 standard, outputting carbon emission intensity indicators for Scope 1-3. As a core component of the platform layer, the carbon accounting engine, based on the international standard ISO 14064, builds dynamic calculation models capable of comprehensive and accurate quantitative analysis of carbon emissions. This engine not only supports carbon emission data acquisition and processing for Scope 1 (direct emissions), Scope 2 (energy indirect emissions), and Scope 3 (other indirect emissions), but also allows for parameterized configuration based on different equipment characteristics, production processes, and energy structures, achieving customized carbon emission intensity indicator output. The carbon accounting engine, combined with a heterogeneous sensor network for real-time acquisition of sensor data, can dynamically update emission factors and activity data, ensuring the timeliness and accuracy of the accounting results.

[0045] The AI ​​analytics module is a crucial component of the platform for intelligent carbon management. It integrates multiple algorithms, including LSTM prediction models, random forest root cause analysis models, and genetic algorithm optimizers, aiming to accurately predict carbon emission trends and generate scientifically feasible emission reduction strategies based on the prediction results. Through the collaborative work of multiple models, this module not only enhances the depth and breadth of data analysis but also provides dynamic optimization support for achieving carbon neutrality goals.

[0046] The LSTM (Long Short-Term Memory) prediction model, a neural network structure specifically designed for processing time-series data, is used in this system for short-term carbon emission trend prediction. The model's input layer receives multi-dimensional time-series data, including historical carbon emission data from the past 24 hours, real-time weather information, and construction schedules. This information collectively constitutes key factors influencing future carbon emission trends. The core of the model consists of a hidden layer containing 128 LSTM units. This layer uses a specific tanh activation function to capture long-term dependencies in the input data, thereby improving prediction accuracy. The output layer generates carbon emission predictions for the next 6, 12, or 24 hours, providing businesses with flexible time window options.

[0047] To maintain the continued effectiveness of the model's predictive power, the LSTM model performs an incremental learning process every two hours, updating the model's weight parameters using the latest collected data. This mechanism enables the model to adapt to changes in the external environment, such as sudden weather changes or adjustments to production plans, thereby ensuring that the prediction results always maintain high accuracy and timeliness.

[0048] In this embodiment, the random forest root cause analysis model is a machine learning method based on decision tree ensemble, primarily used to identify key factors influencing carbon emissions and their mechanisms of action. By constructing multiple decision trees and evaluating feature importance, this model can automatically filter out factors that dominate carbon emission fluctuations from a large number of input variables, such as energy consumption type, equipment operating status, or ambient temperature. Simultaneously, it can reveal the nonlinear relationships and interaction effects between different variables, thus providing a scientific basis for accurately identifying high-carbon emission links and formulating targeted emission reduction measures.

[0049] In terms of generating emission reduction strategies, the system introduces a genetic algorithm optimizer to find the optimal machine scheduling scheme in a manner that simulates the natural evolutionary process. This optimization process begins with initializing the population, generating several sets of candidate machine scheduling schemes, each detailing the start-up and shutdown schedules for various types of machinery. Then, in the fitness calculation phase, the system uses total carbon emissions as the objective function to evaluate the emission reduction effect of each scheduling scheme. To retain high-quality individuals, a roulette wheel selection algorithm is used to select the top 30% of schemes by fitness to participate in subsequent evolutionary operations.

[0050] Next, the system performs crossover and mutation operations on the selected individuals. Single-point crossover generates a new combination scheme with a 70% probability, while random mutation introduces a small perturbation with a 10% probability to prevent the algorithm from getting trapped in local optima. This evolutionary process continues iteratively until a termination condition is met: when the rate of change of the optimal solution over five consecutive generations is less than 1%, the system recognizes the current optimal solution as the final mechanical scheduling scheme and outputs it. In this way, the AI ​​analysis module can automatically generate low-carbon and efficient operating strategies under complex constraints, helping enterprises achieve their green and sustainable development goals.

[0051] As a preferred embodiment, the platform layer further integrates a digital twin simulator to achieve visualized modeling and dynamic simulation of carbon emissions throughout the construction process. This simulator consists of three core units: a BIM model parsing unit, a carbon emission mapping unit, and a scheme comparison unit. These units work collaboratively to construct a virtual-real integrated carbon emission analysis environment.

[0052] The BIM model parsing unit is responsible for interfacing with the Building Information Modeling (BIM) system, automatically extracting the material properties, geometric information, and logical relationships between construction procedures for each component in the model. This data provides a spatial structural foundation and construction process basis for subsequent carbon emission calculations, enabling carbon accounting to not only have a time dimension but also a detailed description at the spatial and technological levels.

[0053] The carbon emission mapping unit spatially binds real-time collected monitoring data on energy consumption, equipment operation, and construction progress to specific components in the BIM model. By introducing a time dimension, a 4D carbon emission distribution model is formed, integrating three-dimensional space and one-dimensional time. This model can dynamically display carbon emission changes at different construction stages and in different areas, improving the spatial interpretability and management granularity of carbon data.

[0054] Based on the aforementioned 4D model, the scheme comparison unit supports virtual simulation of various construction organization schemes and quantifies the impact of different construction sequences or resource allocation methods on total carbon emissions. This unit not only generates intuitive comparison charts but also outputs the carbon emission differences between schemes, assisting project owners in selecting low-carbon and efficient construction strategies during the decision-making phase, thereby achieving carbon optimization management throughout the entire process from design to execution.

[0055] Finally, as a preferred embodiment, the application layer is configured with a large-screen visualization terminal, a mobile app, and a government regulatory API interface, wherein: Visualized large-screen terminals are used for the dynamic presentation of information and data.

[0056] The mobile app includes a foreman terminal and a management terminal. The foreman terminal is used to view machinery energy efficiency rankings and display the carbon emission index corresponding to each unit of workload. It also supports real-time reception of machinery idling warnings sent by edge computing nodes, such as triggering when the power load rate is less than 15% for 10 consecutive minutes. The management terminal is used to present a carbon emission comparison view of multiple projects, generate and present deviation reports from construction contracts, and also supports sorting by dimensions such as unit area and unit output value in the comparison view, as well as analyzing the achievement rate of emission reduction KPIs.

[0057] The government regulatory API interface is used to encapsulate system data in JSON-LD format for government devices to call for supervision. The encapsulated system data fields include basic project information, carbon emission data, and audit trail information. The government regulatory API interface adopts the OAuth 2.0 authorization protocol.

[0058] The project's basic information includes the construction permit number, building area, and the construction company's unified social credit code; carbon emission data is divided into daily cumulative values ​​for Scope 1 / 2 / 3 and the proportion of major emission sources; audit tracking information includes data collection timestamps, edge computing node IDs, and hash verification codes. The application of the OAuth 2.0 authorization protocol allows government regulatory departments to retrieve data in batches by region / time range.

Claims

1. A real-time monitoring and intelligent analysis system for carbon emissions from building construction based on the Internet of Things, characterized in that, The system includes: The perception layer is used to collect multi-dimensional raw data related to carbon emissions at the construction site in real time, including environmental parameters, energy consumption parameters, equipment operating parameters and material flow information, and to perform edge-side preprocessing on the raw data. The transport layer is used to encrypt and transmit pre-processed perception layer data to the platform layer using an adaptive multimode communication protocol, and to adapt to changes in the network environment at the construction site. The platform layer is used to receive and integrate data from the transport layer, perform carbon emission accounting based on real-time data, use artificial intelligence technology to conduct multi-dimensional intelligent analysis of the accounting results and related data, identify emission characteristics, anomalies and optimization potential, and build and update digital twin simulation models of the construction site based on the integrated data to drive the intelligent analysis process. The application layer is used to visualize and dynamically present the carbon emission accounting results, intelligent analysis conclusions, and digital twin simulation status information from the platform layer; it provides interactive access and early warning through mobile applications; and it opens up key carbon emission data and regulatory support to government regulatory platforms through standardized interfaces.

2. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 1, characterized in that, The perception layer is configured with a heterogeneous sensor network, including: Energy consumption sensor arrays are used to collect device-level energy consumption data; Environmental sensor arrays are used to monitor combustion emissions and dust data; Positioning sensor arrays are used to track material transport paths.

3. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 2, characterized in that, The perception layer is also equipped with edge computing nodes and BIM system data interfaces; The heterogeneous sensor network also includes a material consumption monitoring unit, which is used to obtain the amount of various types of building materials entering the site and the actual consumption in real time through the coordinated operation of RFID tags and weighing sensors. The material consumption monitoring unit is connected to the data interface of the BIM system. The material consumption monitoring unit is also used to compare and analyze the deviation between the actual consumption and the planned consumption in the model of the BIM system. When the deviation value exceeds the preset threshold, an early warning signal is triggered. Edge computing nodes have built-in lightweight CNN models, which are used to identify operating conditions and filter outliers from raw data of multiple types of sensors from heterogeneous sensor networks.

4. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 3, characterized in that, Lightweight CNN models include: The input layer is used to receive noise spectrum diagrams and vibration time-series data from raw data of various types of sensors. Convolutional layers are used to extract device operating condition features using kernels of a preset size, and the convolution stride is preset. The pooling layer is used to control the dimensionality reduction ratio to a preset value using the max pooling method. The output layer is used to classify and output the operating status of the equipment.

5. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 1, characterized in that, The multimode communication protocols used in the transport layer include: In pre-designed areas with high concentrations of construction machinery, a self-organizing network is constructed using the LoRa protocol; For mobile devices, the NB-IoT protocol is used, and their data packets are compressed using the COAP protocol.

6. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 5, characterized in that: The transport layer is equipped with a device authentication module, which is used to authenticate each sensor device in the heterogeneous sensor network using a two-way authentication mechanism based on the national cryptographic SM2 algorithm; each sensor device is equipped with a unique digital certificate.

7. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 1, characterized in that, The platform layer is equipped with a carbon accounting engine and an AI analysis module, including: The carbon accounting engine is used to build dynamic calculation models based on the ISO 14064 standard and output carbon emission intensity indicators in Scope 1-3. The AI ​​analysis module integrates an LSTM prediction model, a random forest root cause analysis model, and a genetic algorithm optimizer. It is used to predict carbon emission trends and generate emission reduction strategies.

8. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 7, characterized in that, LSTM prediction models include: The input layer is used to receive carbon emission time-series data, meteorological data, and construction progress data for a preset historical period. The hidden layer has a preset number of LSTM units and corresponding activation functions. The hidden layer is used to receive various data inputs from the input layer, calculate them, and output them to the output layer. The output layer is used to generate carbon emission predictions for a preset future time period. The LSTM prediction model is also used to perform incremental learning algorithms within a preset period to update its own model weight parameters; The genetic algorithm optimizer is used to output the optimal scheduling scheme by executing the execution logic; the execution logic is as follows: Initialize the population: Generate a preset number of machine scheduling schemes, each of which contains the start and stop schedules for each type of machine; Fitness calculation: Using the minimization of total carbon emissions as the objective function, calculate the fitness value for each group of machine scheduling schemes; Selection operation: Use roulette wheel selection to retain individuals with the highest fitness ranking in the first preset percentage; Crossover mutation: Perform single-point crossover and random mutation with preset probabilities on the retained individuals; Iteration termination condition: After a preset number of iterations, the optimal solution is output as the optimal scheduling scheme when the rate of change of the optimal solution is less than a second preset percentage.

9. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 7, characterized in that, The platform layer is also equipped with a digital twin emulator, including: BIM model parsing unit is used to extract the component material properties and construction sequence logical relationships of the model in the BIM system; The carbon emission mapping unit is used to spatially bind real-time collected and monitored data with model components in the BIM system to form a 4D carbon emission distribution model. The scheme comparison unit is used to simulate the impact of different construction sequences on total carbon emissions and calculate and quantify the differences between the schemes.

10. The real-time monitoring and intelligent analysis system for carbon emissions from building construction according to claim 1, characterized in that, The application layer is configured with a large-screen visualization terminal, a mobile app, and government regulatory API interfaces, including: Large-screen visualization terminals are used for the dynamic and visual presentation of information and data. The mobile app includes a foreman terminal and a management terminal. The foreman terminal is used to view the mechanical energy efficiency ranking and display the carbon emission index corresponding to the unit workload. The management terminal is used to present a comparison view of carbon emissions from multiple projects, generate and present deviation reports from the construction contract. The government regulatory API interface is used to encapsulate system data in JSON-LD format for government devices to call for supervision. The encapsulated system data fields include basic project information, carbon emission data, and audit trail information. The government regulatory API interface adopts the OAuth 2.0 authorization protocol.

Images