Green building design decoration engineering management method and system

Abstract

The invention relates to the technical field of building engineering management, in particular to a green building design and decoration engineering management method and system. The system comprises a cross-modal data fusion engine, a dynamic energy consumption-carbon emission coupling prediction module, an intelligent material selection and supply chain cooperation module, a block chain quality tracing module and a visual decision terminal. The data collection module collects BIM, GIS and other multi-source data, a standardized data set is output through fusion of an improved attention mechanism, the prediction module predicts energy consumption and carbon emission through a bidirectional LSTM-GBRT model, the material supply chain module selects materials through a multi-objective optimization algorithm and generates a distribution plan, the block chain module stores inspection and other data, and the terminal assists in decision making. The method comprises five steps of data fusion, carbon consumption prediction, green decision making, process management and control and full-period tracing. And full-cycle green management and control of the project can be realized, data cannot be tampered, and efficient green decision is supported.

Description

Technical Field

[0001] This invention relates to the field of building engineering management technology, specifically a method and system for managing green building design and decoration projects. Background Technology

[0002] With the development of the global low-carbon economy and the proposal of "dual-carbon" goals, green building has become the mainstream of the construction industry. Green building design and decoration engineering emphasizes saving resources, protecting the environment, and reducing pollution throughout the entire life cycle, providing people with healthy, suitable, and efficient living spaces. However, there are still many problems to be solved in the current management of green building design and decoration engineering, and existing technologies are insufficient to meet the management needs of the entire life cycle, high precision, and green development.

[0003] First, existing engineering management systems suffer from a disconnect between design and construction. In traditional management models, BIM model data and green indicators from the design phase are often difficult to effectively transfer to the construction and finishing phases. Construction units primarily rely on two-dimensional drawings for construction, failing to fully utilize the three-dimensional data and green parameters from the design phase. For example, while BIM technology is introduced, it is often used only for construction progress simulation, without achieving real-time linkage between green indicators from the design phase and the construction process. This makes it difficult to implement green concepts from the design phase during construction, leading to problems such as material selection not meeting design requirements and construction techniques failing to meet green energy consumption standards.

[0004] Secondly, the management of green indicators is lagging and lacks accuracy. Current technologies for monitoring engineering energy consumption and carbon emissions mostly rely on post-construction statistics, meaning calculations are made after project completion based on construction records and material consumption data, making real-time adjustments during construction impossible. Even systems that incorporate real-time monitoring modules often use single linear prediction models, failing to consider the coupled effects of multiple factors such as environmental parameters (e.g., temperature, humidity, wind speed), construction progress, and material properties, leading to significant prediction deviations. For example, during high-temperature construction in summer, the energy consumption of air conditioning and construction machinery increases significantly. If predictions are based solely on conventional operating conditions, actual energy consumption and carbon emissions will be underestimated, making it impossible to implement timely energy-saving control measures and impacting the project's green compliance rate.

[0005] Secondly, the selection of green materials lacks scientific basis and suffers from insufficient supply chain coordination. Green buildings have high requirements for the environmental performance of materials, requiring comprehensive consideration of multiple indicators such as carbon emissions, formaldehyde release, and recyclability throughout the material's life cycle. In current technologies, material selection largely relies on the experience and judgment of management personnel, lacking systematic integration and analysis of material life cycle data (LCA data). Furthermore, the disconnect between material procurement and supply chain management makes it impossible to monitor supplier capacity, delivery cycles, and material inventory status in real time, easily leading to material supply delays or stockpiles. This not only affects the construction period but may also result in the abandonment of green materials due to emergency procurement, lowering the project's green rating.

[0006] Furthermore, the traceability system for project quality and green indicators is inadequate. Existing quality traceability systems mostly rely on paper records or centralized databases, making data susceptible to tampering and loss, and failing to guarantee the authenticity and completeness of traceability information. When green indicators fail to meet standards or quality issues arise, it is difficult to quickly pinpoint the root cause, such as accurately tracing the source of substandard materials or the responsible party for construction procedures, increasing the difficulty and cost of rectification. Simultaneously, existing traceability systems are mostly limited to the construction phase, failing to achieve full-cycle traceability from design and material procurement to completion and operation, thus failing to meet the requirements of green building lifecycle management.

[0007] Finally, existing technologies lack sufficient multi-source data integration capabilities. Green building design and decoration engineering involves complex data types, including BIM 3D data, GIS geospatial data, IoT real-time monitoring data, and material LCA data, etc. These data sources are scattered and their formats are inconsistent. Existing systems lack effective data fusion technologies, resulting in low data utilization and an inability to provide comprehensive and accurate support for management decisions. For example, GIS data is used for site planning but cannot be effectively combined with construction progress data from the BIM model, making it impossible to optimize construction plans based on the site environment; IoT monitoring data is disconnected from material energy consumption data, making it impossible to accurately analyze the energy consumption performance of different materials under actual working conditions.

[0008] In summary, existing green building design and decoration engineering management technologies suffer from several shortcomings, including a disconnect between design and construction, lagging control of green indicators, unscientific material selection, insufficient supply chain collaboration, an imperfect traceability system, and weak data integration capabilities. These deficiencies make it difficult to meet the full-cycle, high-precision, and green management needs of green buildings, thus hindering the healthy development of the green building industry. Therefore, developing a management system and method that enables integrated full-cycle management, real-time and precise control of green indicators, scientific selection of green materials, and guaranteed data traceability has significant practical implications and application value. Summary of the Invention

[0009] To address the shortcomings of existing technologies, such as the disconnect between design and construction, lagging green indicator control, unscientific material selection, insufficient supply chain collaboration, imperfect traceability system, and weak data integration capabilities, this invention provides a green building design and decoration engineering management system and method. This system enables integrated management of the entire lifecycle of green building design and decoration engineering, improves the accuracy of green indicator control, ensures the green compliance rate of projects, optimizes the construction period and cost, and improves project management efficiency.

[0010] The technical solution adopted by this invention to solve its technical problem is: a green building design and decoration engineering management system, comprising: a cross-modal data fusion engine, used to collect architectural design BIM data, site GIS data, environmental IoT sensor data, and green material LCA full life cycle data, and adopts an improved attention mechanism fusion algorithm to clean, align, and fuse multi-source heterogeneous data, outputting a standardized fusion dataset; a dynamic energy consumption-carbon emission coupled prediction module, which communicates with the cross-modal data fusion engine, constructs a bidirectional LSTM-GBRT coupled model based on the fusion dataset, inputs construction schedule, environmental parameters, and material usage scheme, and outputs real-time prediction results of energy consumption and carbon emissions at each stage of the project; and intelligent material selection and supply chain management. The collaboration module connects to the cross-modal data fusion engine and the dynamic energy consumption-carbon emission coupled prediction module, respectively. It has a built-in green material database and a real-time supply chain status monitoring unit. Based on the prediction results and engineering design requirements, it selects the optimal combination of green materials through a multi-objective optimization algorithm and generates a supply chain distribution plan. The blockchain quality traceability module connects to the intelligent material selection and supply chain collaboration module and the construction monitoring unit. It adopts a consortium blockchain architecture and writes material arrival inspection data, construction process quality data, and green indicator compliance data into the blockchain node to achieve tamper-proof data traceability. The visualization decision terminal communicates with the above modules to display fused data, prediction results, material selection schemes, and traceability information, and supports managers to issue control instructions.

[0011] Specifically, the cross-modal data fusion engine includes a data acquisition unit, a data preprocessing unit, and a fusion computing unit; the data acquisition unit is equipped with a BIM model parsing interface, a GIS map calling interface, an IoT sensor gateway, and an LCA data docking interface, supporting the simultaneous acquisition of three-dimensional geometric data, geospatial data, real-time environmental data, and material life cycle data.

[0012] Specifically, the bidirectional LSTM-GBRT coupling model in the dynamic energy consumption-carbon emission coupled prediction module uses a bidirectional LSTM network to extract the correlation features between construction progress and energy consumption in the time dimension, and uses the GBRT model to fit the nonlinear relationship between environmental parameters and carbon emissions. The output results of the two are then weighted and fused to obtain the final prediction value.

[0013] Specifically, the multi-objective optimization algorithm of the intelligent material selection and supply chain collaboration module aims to minimize carbon emissions, optimize costs, and shorten the construction period. The constraints include material performance compliance, supply chain delivery capabilities, and on-site storage capacity.

[0014] Specifically, the blockchain quality traceability module includes a data acquisition terminal, a consortium blockchain node server, and a traceability query unit; the consortium blockchain nodes include construction unit nodes, construction unit nodes, supervision unit nodes, and material supplier nodes, and each node uses the PBFT consensus mechanism to achieve data consensus verification.

[0015] A green building design and decoration engineering management method based on the above system includes the following steps: S1. Data Acquisition and Fusion: The cross-modal data fusion engine collects architectural design BIM data, site GIS data, environmental IoT sensor data, and green material LCA data. After preprocessing, an improved attention mechanism fusion algorithm is used to generate a standardized fusion dataset. S2. Dynamic carbon consumption prediction: Input the fused dataset and construction schedule into the bidirectional LSTM-GBRT coupled model to predict the energy consumption and carbon emissions at each stage of the project in real time. If the predicted value exceeds the preset green threshold, an early warning will be triggered. S3, Green Decision Generation: The intelligent material selection and supply chain collaboration module uses a multi-objective optimization algorithm to select green material combinations and generate material procurement and distribution plans based on prediction results and design requirements. S4. Construction process control: During the construction process, process quality data and green indicator compliance data are collected in real time and uploaded to the blockchain quality traceability module. Managers monitor the construction status through a visual decision terminal and adjust the construction plan according to the early warning information. S5. Full-cycle traceability: After the project is completed, the blockchain quality traceability module enables full-cycle traceability and query of material sources, construction procedures, and compliance with green indicators.

[0016] Specifically, the improved attention mechanism fusion algorithm in step S1 calculates the information entropy weights of each modality data and performs weighted fusion of data from different sources. The weight update cycle is consistent with the IoT data collection cycle.

[0017] Specifically, in step S2, if the predicted value exceeds the preset threshold, the triggered warning information includes the type of exceedance, the extent of exceedance, and suggested adjustment schemes. Suggested adjustment schemes include material replacement suggestions and construction process optimization suggestions.

[0018] Specifically, the construction process control in step S4 also includes real-time monitoring of the energy consumption of construction machinery, collecting machinery operating parameters through IoT sensor nodes, calculating the deviation between actual energy consumption and predicted energy consumption, and adjusting the machinery usage plan in a timely manner.

[0019] Specifically, the full-cycle traceability query in step S5 supports multi-dimensional queries based on engineering stage, material type, and quality inspection items, and the query results generate a traceability report with timestamps.

[0020] The beneficial effects of this invention are: 1. Achieving deep integration of design and construction: By integrating BIM data from the design phase with real-time data from the construction phase through a cross-modal data fusion engine, this ensures that green indicators from the design phase are effectively transmitted to the construction process, solving the problem of disconnect between design and construction in existing technologies. Implementation shows that this invention can increase the implementation rate of green design indicators in construction to over 95%.

[0021] 2. Improve the accuracy and real-time performance of green indicator management: Construct a bidirectional LSTM-GBRT coupled prediction model, considering the coupling effect of multiple factors, and improve the prediction accuracy compared with the existing linear model; realize real-time prediction and early warning of energy consumption and carbon emissions, and adjust the construction plan in a timely manner to avoid post-event rectification and improve the green compliance rate of the project.

[0022] 3. Scientifically optimize the selection of green materials and supply chain collaboration: Based on multi-objective optimization algorithms and material LCA data, the selection is carried out to ensure that the green performance of materials meets the standards. At the same time, the delivery plan is generated in combination with the real-time status of the supply chain to reduce material supply delays, shorten the construction period, and reduce material procurement costs.

[0023] 4. Construct an immutable full-lifecycle traceability system: Adopt a consortium blockchain architecture to achieve data storage and ensure the authenticity and integrity of traceability information. This can quickly locate the root cause of problems, reduce the cost of problem rectification, and meet the requirements of green building full lifecycle management.

[0024] 5. Improve the efficiency of multi-source data integration and utilization: The improved attention mechanism fusion algorithm can effectively integrate heterogeneous data, improve data utilization compared with the existing system, provide comprehensive and accurate support for management decisions, and improve engineering management efficiency. Attached Figure Description

[0025] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] Figure 1 An architecture diagram of a green building design and decoration engineering management system provided by the present invention; Figure 2 A flowchart of a green building design and decoration engineering management method provided by the present invention. Detailed Implementation

[0027] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0028] like Figure 1 As shown, the green building design and decoration engineering management system of this invention includes a cross-modal data fusion engine, a dynamic energy consumption-carbon emission coupled prediction module, an intelligent material selection and supply chain collaboration module, a blockchain quality traceability module, and a visualization decision-making terminal. Each module interacts with the other via industrial Ethernet and wireless communication modules. The specific structure is as follows: Cross-modal data fusion engine: As the data core of the system, it is responsible for the collection, preprocessing, and fusion of multi-source heterogeneous data. The data acquisition unit is equipped with a BIM model parsing interface (supporting mainstream BIM software formats such as Revit and Bentley), a GIS map call interface (connecting to Gaode Map and Baidu Map API), an IoT sensor gateway (supporting LoRa and NB-IoT communication protocols), and an LCA data interface (connecting to the international EPD database and the domestic building materials LCA database). It can simultaneously collect 3D geometric data of building design, site geospatial data (topography, traffic, surrounding environment), real-time environmental data (temperature, humidity, wind speed, PM2.5), construction machinery operating parameters (speed, power, fuel consumption), and green material LCA data (lifecycle carbon emissions, formaldehyde release, recyclability, production cost). The data preprocessing unit uses outlier detection (based on the 3σ criterion), missing value imputation (based on the KNN algorithm), and data standardization (min-max normalization) techniques to process the collected data. The fusion computing unit uses an improved attention mechanism fusion algorithm to calculate the information entropy weights of each modality (the higher the information entropy, the greater the weight) and perform weighted fusion of data from different sources to generate a standardized fusion dataset. The weight update cycle is consistent with the IoT data collection cycle (10 minutes / time) to ensure the real-time performance and accuracy of the fused data.

[0029] The dynamic energy consumption-carbon emission coupled prediction module communicates with the cross-modal data fusion engine via TCP / IP protocol and constructs a bidirectional LSTM-GBRT coupled prediction model based on the fused dataset. The bidirectional LSTM network is responsible for extracting the correlation features in the time dimension, namely the temporal correlation between construction progress (percentage of completion of each process) and energy consumption and carbon emissions. Its input layer is the construction progress data and mechanical operation parameters in the fused dataset, with two hidden layers, each with 64 neurons, and the output layer is the preliminary energy consumption prediction value. The GBRT model is responsible for fitting the nonlinear relationship. Its input layer is the environmental parameters and material performance data in the fused dataset. By constructing multiple regression trees, it fits the nonlinear correlation between environmental parameters and carbon emissions and outputs the preliminary carbon emission prediction value. The coupled model fuses the output results of the two through adaptive weighting coefficients (dynamically adjusted based on prediction error) to obtain the final predicted energy consumption and carbon emissions for each stage of the project (design, foundation construction, main structure construction, decoration construction, and completion). The module has a built-in green threshold database (based on GB / T 50378-2019 "Evaluation Standard for Green Buildings"). If the predicted value exceeds the preset threshold of the corresponding stage, an early warning will be triggered immediately, generating warning information that includes the type of exceedance (exceeding energy consumption standards / exceeding carbon emission standards), the extent of exceedance, and adjustment suggestions (material replacement, process optimization, and machinery scheduling).

[0030] The intelligent material selection and supply chain collaboration module adopts a distributed architecture, connecting to the cross-modal data fusion engine and the dynamic energy consumption-carbon emission coupled prediction module via message queues (RabbitMQ). The module has a built-in green materials database, storing LCA data, performance parameters (strength, durability, fire rating), prices, and supplier information for various building and decorative materials. The real-time supply chain status monitoring unit connects to supplier ERP systems and logistics tracking systems to collect real-time data on supplier capacity, inventory, delivery vehicle location, and transportation time. Based on the prediction results from the dynamic energy consumption-carbon emission coupled prediction module and engineering design requirements (such as green building star rating standards and functional requirements), the module selects materials using a multi-objective optimization algorithm (NSGA-Ⅲ algorithm). The optimization objectives are lowest carbon emissions, optimal cost, and shortest construction period. Constraints include material performance compliance, supply chain delivery capacity (delivery cycle ≤ 7 days), and on-site storage capacity (≤ 500m³). 2 Once the selection is complete, an optimal list of green materials combinations is generated, and a delivery plan is created based on the real-time status of the supply chain, including delivery time, transportation routes, unloading locations, and storage solutions. The plan is then pushed to suppliers and construction units.

[0031] The blockchain quality traceability module adopts a consortium blockchain architecture, with nodes including those of the construction unit, contractor, supervision unit, and material suppliers. These nodes are interconnected via a P2P network and utilize the PBFT consensus mechanism (consensus latency ≤ 3 seconds) for data consensus verification. The module is equipped with data acquisition terminals (including RFID readers, high-definition cameras, and handheld testing devices) to collect data on incoming material inspections (material type, batch, LCA test report, appearance quality), construction process quality data (process completion time, construction personnel, testing indicators), and green indicator compliance data (actual energy consumption, carbon emissions, indoor air quality). The collected data is written to the blockchain after consensus verification by each node, generating timestamped blocks to ensure immutable data storage. The module also includes a traceability query unit, supporting managers to perform multi-dimensional queries using keywords such as material number, process number, and time, generating traceability reports and clearly defining the responsible parties at each stage.

[0032] Visualized Decision Terminal: Employing an industrial-grade touchscreen display, this terminal utilizes WebGL technology to construct a 3D visualization interface that communicates with each of the aforementioned modules. It can display standardized fusion datasets (overlaying BIM models and GIS maps, real-time environmental data curves), dynamic energy consumption and carbon emission prediction results (bar charts, line graphs), intelligent material selection solutions (material lists, performance parameter comparisons), supply chain delivery progress (logistics trajectory), and blockchain traceability information in real time. Managers can issue control commands through the terminal, such as adjusting material procurement plans, optimizing construction schedules, and activating energy-saving measures. These commands are encrypted and transmitted to the corresponding modules for execution.

[0033] like Figure 2 As shown, the green building design and decoration engineering management method of the present invention includes the following steps: S1. Data Acquisition and Fusion: The cross-modal data fusion engine is activated, collecting BIM data (building structure, unit layout, decorative details), site GIS data (site topography, surrounding roads, municipal pipelines), environmental IoT sensor data (real-time temperature, humidity, wind speed), construction machinery operating parameters (engine speed, actual power, fuel consumption), and green material LCA data (lifecycle carbon emissions and recyclability of materials such as concrete, steel bars, and coatings obtained from the EPD database) through various interfaces and sensing devices. The data preprocessing unit uses the 3σ criterion to remove outliers in the environmental data, fills missing values ​​in the material LCA data using the KNN algorithm, and standardizes each data point to the [0,1] range using min-max normalization. The fusion calculation unit uses an improved attention mechanism fusion algorithm to calculate the information entropy weights of each modality (e.g., environmental data information entropy 0.32, BIM data 0.28, LCA data 0.25, machinery parameters 0.15), performs weighted fusion on the standardized data, and generates a standardized fusion dataset, which is updated every 10 minutes.

[0034] S2. Dynamic Carbon Consumption Prediction: The standardized fusion dataset generated in step S1 and the construction schedule (e.g., foundation construction days 1-10, main structure construction days 11-40, decoration construction days 41-60) are input into the dynamic energy consumption-carbon emission coupled prediction module. A bidirectional LSTM network extracts the temporal correlation features between construction progress and energy consumption, outputting preliminary energy consumption predictions. The GBRT model fits the nonlinear relationship between environmental parameters and carbon emissions, outputting preliminary carbon emission predictions. The coupled model uses adaptive weighting coefficients (e.g., energy consumption prediction weight 0.45, carbon emission prediction weight 0.55) to fuse the predictions and obtain the final prediction results, generating energy consumption and carbon emission prediction curves for each construction stage. The module compares the prediction results with a preset green threshold. If the predicted carbon emissions for a certain stage exceed the threshold (e.g., the threshold for the decoration construction stage is 50 kg CO2 / m³), the prediction is triggered. 2 If the carbon emission during the decoration construction phase is predicted to be 58 kg CO2 / m³, an early warning will be triggered, generating the warning message: "Predicted carbon emission during the decoration construction phase: 58 kg CO2 / m³". 2 The emissions exceeded the standard by 16%, and it is recommended to replace the coatings with low-emission coatings (replacing traditional solvent-based coatings with water-based coatings) and optimize the ventilation process.

[0035] S3. Green Decision Generation: The intelligent material selection and supply chain collaboration module receives early warning information and the fused dataset from step S1, calls the built-in green material database, and filters out low-emission coatings that meet the requirements of decoration construction (such as water-based acrylic coatings, whose life cycle carbon emissions are reduced by 35% compared to traditional coatings). Multi-objective optimization is performed using the NSGA-Ⅲ algorithm, with input optimization objectives (lowest carbon emissions, optimal cost, shortest construction period) and constraints (coating fire resistance rating ≥ Class A, delivery cycle ≤ 5 days, on-site storage capacity ≤ 100m³). 2After 50 iterations, the algorithm outputs an optimal material combination list (water-based acrylic paint, environmentally friendly putty, and biodegradable wallpaper). Simultaneously, the supply chain real-time status monitoring unit connects to the paint supplier's ERP system to obtain supplier inventory (1000 barrels of water-based acrylic paint, meeting demand) and delivery vehicle location (30km from the construction site). A delivery plan is generated: "At 9:00 AM on [Date], 202X, a delivery vehicle with license plate number XXX will deliver 100 barrels of water-based acrylic paint to the storage warehouse in Area A of the construction site, equipped with moisture-proof storage facilities," and the plan is then pushed to the supplier and construction unit.

[0036] S4. Construction Process Control: The construction unit receives materials according to the material delivery plan and reads the RFID tags on the material packaging through the RFID reader of the blockchain quality traceability module, collecting information such as material model, batch, and LCA test report. The supervision unit verifies the information, and writes it into the blockchain after successful verification. During construction, IoT sensing devices collect real-time operating parameters of construction machinery and environmental data, uploading them to the cross-modal data fusion engine to update the fusion dataset. The dynamic energy consumption-carbon emission coupled prediction module adjusts the prediction results in real time based on the updated dataset. Management personnel monitor the construction progress and green indicator compliance through a visual decision terminal. If the actual energy consumption of a certain process exceeds the predicted value by 5%, a control instruction is issued: "Adjust the operating parameters of the construction machinery, reduce the excavator speed from 2200r / min to 2000r / min, and reduce idling time." After the construction unit executes the instruction, the module re-predicts the energy consumption to ensure that the energy consumption is controlled within the threshold range. At the same time, quality inspection personnel collect process quality data (such as paint thickness and smoothness) through handheld inspection devices and upload them to the blockchain quality traceability module to achieve real-time control of the construction process.

[0037] S5. Full-Cycle Traceability: After project completion, the blockchain quality traceability module stores full-cycle information from BIM data and material procurement information in the design phase to quality data and green indicator compliance data during construction. Managers can use the traceability query unit to input the material number "SC202X0508" to query the source (supplier name, production time), incoming inspection data (inspection time, inspector, pass / fail status), application location (living room wall, bedroom wall), and corresponding green indicator (indoor formaldehyde concentration after painting: 0.03 mg / m³). 3 (Compliant with GB 50325-2020 standard); Enter the process number "GZ202X0615" to query the construction personnel, construction time, quality inspection results, and supervision opinions for this decoration process. The query results generate a traceability report with a timestamp, which can be used as a basis for green building evaluation, project quality acceptance, and subsequent operation and maintenance.

[0038] Example Example 1: Management of Design and Decoration Projects for New Green Residential Buildings This example is applied to a newly built green residential project with a total construction area of ​​20,000 m². 2 The project comprises 10 buildings, each 18 stories high, designed to meet the two-star green building standard. The green building design and decoration engineering management system and method of this invention are used for full-cycle management. The specific implementation process is as follows: 1. System Deployment: A cross-modal data fusion engine server (configured with an Intel Xeon E5-2680 v4 processor, 32GB of memory, and a 2TB hard drive) is set up at the project construction site. A BIM model parsing interface (interfacing with Revit 2023 software) and a GIS map call interface (interfacing with Gaode Map API) are deployed. IoT sensing devices (including 10 temperature and humidity sensors, 5 wind speed sensors, and 20 construction machinery operation parameter acquisition terminals) are installed, and sensor data transmission is achieved through a LoRa gateway. The intelligent material selection and supply chain collaboration module interfaces with the domestic building materials LCA database and the ERP systems of three major material suppliers (concrete supplier, paint supplier, and tile supplier). The blockchain quality traceability module deploys four consortium blockchain nodes (one each for the construction unit, contractor, supervision unit, and material supplier), each configured with a Lenovo ThinkSystem SR650 server, using the PBFT consensus mechanism. Visual decision-making terminals are set up at the project department and the construction unit's office at the construction site, using 27-inch industrial touch screens and a 3D visualization interface built based on WebGL technology.

[0039] 2. Data Acquisition and Fusion: The cross-modal data fusion engine is activated. Project BIM model data, including 3D geometric data such as building structure, unit layout, and decorative details (e.g., wall decoration, floor decoration, ceiling decoration), is imported through the BIM model parsing interface. Site GIS data, including site topography (5° slope), surrounding roads (300m from main roads), and municipal pipelines (location of water, electricity, and gas pipelines), is obtained through the GIS map call interface. Real-time environmental data (average temperature 28℃, humidity 65%, wind speed 2.5m / s during construction) and construction machinery operating parameters (excavator speed 2200r / min, power 150kW, fuel consumption 25L / h; tower crane speed 15r / min, power 75kW, fuel consumption 12L / h) are collected from the domestic building materials LCA database through the LCA data interface. (Concrete lifecycle carbon emissions: 300kg / m³). 3The recyclability rate is 20%; the life cycle carbon emission of water-based coatings is 50 kg / barrel, and the recyclability rate is 80%. The data preprocessing unit uses the 3σ criterion to remove one set of abnormal temperature data (45℃, exceeding the normal construction temperature range), fills two sets of missing coating LCA data using the KNN algorithm, and uses min-max normalization to standardize each data to the [0,1] interval. The fusion calculation unit uses an improved attention mechanism fusion algorithm to calculate the information entropy weight of each modality data: BIM data 0.28, GIS data 0.20, environmental data 0.22, mechanical parameters 0.15, LCA data 0.15. The standardized data is then weighted and fused to generate a standardized fusion dataset, which is updated every 10 minutes.

[0040] 3. Dynamic Carbon Consumption Prediction: The standardized fusion dataset and construction schedule (20 days for foundation construction, 60 days for main structure construction, 40 days for decoration construction, and 10 days for completion) are input into the dynamic energy consumption-carbon emission coupled prediction module. A bidirectional LSTM network with two hidden layers and 64 neurons per layer is trained iteratively 100 times to extract the temporal correlation features between construction progress and energy consumption, and output preliminary energy consumption predictions (5000 kWh for foundation construction, 15000 kWh for main structure construction, and 8000 kWh for decoration construction). The GBRT model constructs 50 regression trees to fit the nonlinear relationship between environmental parameters and carbon emissions, and outputs preliminary carbon emission predictions (80 tCO2 for foundation construction, 220 tCO2 for main structure construction, and 60 tCO2 for decoration construction). The coupled model uses adaptive weighting coefficients (0.45 for energy consumption prediction and 0.55 for carbon emission prediction) to fuse the final prediction results, generating energy consumption and carbon emission prediction curves for each construction stage. The module compares the predicted results with the preset thresholds of the two-star green building standard (energy consumption threshold of 8500kWh and carbon emission threshold of 55tCO2 during the decoration and construction stage). It finds that the predicted carbon emission value during the decoration and construction stage is 60tCO2, exceeding the standard by 9.1%, triggering an early warning and generating the warning message: "The predicted carbon emission value during the decoration and construction stage is 60tCO2, exceeding the standard by 9.1%. It is recommended to replace the tiles with low-emission tiles (replace ordinary tiles with recycled aggregate tiles) and optimize the ceiling construction process (use light steel keel instead of wooden keel to reduce the use of wood)."

[0041] 4. Green Decision Generation: The intelligent material selection and supply chain collaboration module receives early warning information and fused datasets, calls upon the built-in green material database, and selects recycled aggregate tiles that meet the requirements for decorative construction (lifecycle carbon emissions are reduced by 40% compared to ordinary tiles, strength ≥5MPa, fire resistance rating A). Multi-objective optimization is performed using the NSGA-Ⅲ algorithm, with input optimization objectives (lowest carbon emissions, optimal cost, shortest construction period) and constraints (tile size 600mm×600mm, delivery cycle ≤7 days, on-site storage capacity ≤200m³). 2 After 50 iterations, the algorithm outputs an optimal material combination list: recycled aggregate ceramic tiles, water-based acrylic paint, light steel keel, and environmentally friendly putty. Simultaneously, the supply chain real-time status monitoring unit connects to the ceramic tile supplier's ERP system to obtain supplier inventory (currently 5000m³ of recycled aggregate ceramic tiles). 2 To meet the project's decoration and construction needs of 2000m 2 ), Delivery vehicle location (50km from the construction site), generate delivery plan: "At 2 PM on [Date], a delivery vehicle with license plate number Yue AXXX will deliver 2000m..." 2 The recycled aggregate tiles were delivered to the storage warehouse in Area B of the construction site, which was equipped with rainproof and moisture-proof storage facilities. After unloading, the supervision unit conducted an on-site inspection and then sent the plan to the tile supplier and the construction unit.

[0042] 5. Construction Process Control: The construction unit receives recycled aggregate tiles according to the delivery plan. Using the RFID reader in the blockchain quality traceability module, the RFID tags on the tile packaging are read to collect information such as material model (ZSSG-600×600), batch number (202X0510), and LCA test report (carbon emissions 120kg / m³). 2Information such as material performance (confirming that the material performance meets design requirements) is collected and verified by the supervision unit. Once verified, the information is written to the blockchain, generating a timestamp "202X-05-12 14:30:00". During construction, IoT sensors collect real-time operating parameters of construction machinery (tile laying machine speed 30r / min, power 5kW, fuel consumption 3L / h) and environmental data (real-time temperature 30℃, humidity 60%), uploading them to the cross-modal data fusion engine to update the fused dataset. The dynamic energy consumption-carbon emission coupled prediction module adjusts the prediction results in real-time based on the updated dataset (the predicted carbon emission value for the decoration construction stage is corrected to 53tCO2, lower than the threshold of 55tCO2). Managers monitored the construction progress (wall surface treatment was completed on the 15th day of decoration construction, in line with the schedule) and the compliance of green indicators through a visual decision-making terminal. They discovered a 2cm deviation in the spacing of the light steel keel installation during the ceiling construction process (allowable deviation is 1cm). They issued a control order: "Immediately rectify the installation of the ceiling light steel keel, adjust the spacing to 40cm (design requirement is 40cm±1cm), and re-inspect after rectification." After the construction unit implemented the rectification, quality inspection personnel collected the data after rectification (spacing 39.5cm) using a handheld testing device, uploaded it to the blockchain quality traceability module, and wrote it into the blockchain after verification.

[0043] 6. Full-cycle traceability: After project completion, management personnel can conduct multi-dimensional queries through the traceability query unit of the blockchain quality traceability module. Entering the material number "ZSSG-202X0510" reveals that the supplier of this batch of recycled aggregate tiles is XX New Building Materials Co., Ltd., the production date is May 8, 202X, the on-site inspection date is May 12, 202X, the inspector is XXX, and the status is qualified; the installation locations are living room and bedroom floors, and the corresponding green indicator (indoor formaldehyde concentration after installation is 0.02mg / m³) is... 3 The project meets the GB 50325-2020 standard, reducing carbon emissions by 80tCO2 compared to ordinary ceramic tiles. Entering the process number "ZS-202X0608" reveals the construction worker XXX for this ceiling installation process, the construction date as June 8, 202X, the initial spacing as 42cm, and the subsequent spacing as 39.5cm. The supervisor's opinion is "qualified." The query results generate a timestamped traceability report, serving as a crucial basis for the green building two-star evaluation. The project ultimately passed the green building two-star review successfully.

[0044] In this embodiment, by adopting the system and method of the present invention, the actual carbon emissions during the decoration construction phase were 52tCO2, which is 13.3% lower than the predicted excess value and 22% lower than the traditional construction method; the construction period was shortened by 10 days (8.3%) compared to the plan; material procurement costs were reduced by 7%; and the green indicator compliance rate was 98.5%, achieving efficient and green management of green building design and decoration engineering.

[0045] Example 2: Green Renovation Design and Decoration Engineering Management of Existing Buildings This example is applied to a green renovation project of an existing office building with a building area of ​​8000m². 2 Built in 2010, the renovation aimed to upgrade to a one-star green building standard. The main renovations included exterior wall insulation and decoration, interior renovation, and upgrades to the water supply, drainage, and electrical systems. The green building design and decoration engineering management system and method of this invention were used for full-cycle management. The specific implementation process is as follows: 1. System Deployment: A cross-modal data fusion engine server (configured with an Intel Xeon E5-2678 v3 processor, 24GB RAM, and a 1TB hard drive) and a BIM model parsing interface (connecting to Bentley MicroStation software for importing BIM data from reverse modeling of existing buildings) and a GIS map call interface (connecting to Baidu Maps API) are deployed at the project site. IoT sensing devices (8 temperature and humidity sensors, 3 wind speed sensors, and 15 construction machinery operation parameter acquisition terminals) are installed, and sensor data transmission is achieved through an NB-IoT gateway. The intelligent material selection and supply chain collaboration module connects to the international EPD database and the ERP systems of two renovation material suppliers (exterior wall insulation material supplier and interior decoration material supplier). The blockchain quality traceability module deploys 4 consortium blockchain nodes, adopting the PBFT consensus mechanism. Visual decision-making terminals are set up at the project site's project management office and the construction unit's (office building owner's) office.

[0046] 2. Data Acquisition and Fusion: The cross-modal data fusion engine is activated. Existing building reverse-modeling BIM data is imported via the BIM model parsing interface, including the original building structure (frame structure), existing decorative details (original walls are painted with ordinary latex paint, floors are tiled, ceilings are plasterboard), and equipment pipelines (water supply and drainage pipes, electrical wiring), etc., in 3D geometric data. GIS site GIS data is obtained via the GIS map call interface, including the surrounding environment of the office building (500m from the commercial center), traffic conditions (100m from the nearest bus stop), and municipal facilities. (Power supply capacity, water supply pressure); IoT sensing devices collect environmental data in real time during the renovation period (average temperature 25℃, humidity 70%, wind speed 1.8m / s), and operating parameters of construction machinery (electric drill speed 3000r / min, power 2kW; cutting machine speed 4500r / min, power 3kW); obtain LCA data of renovation materials such as exterior wall insulation rock wool board, water-based latex paint, recycled plastic flooring, and LED lighting from the international EPD database through the LCA data interface (exterior wall insulation rock wool board life cycle carbon emissions 80kg / m³). 2The recyclability rate is 70%; the lifecycle carbon emission of water-based latex paint is 45 kg / barrel, and the recyclability rate is 85%. The data preprocessing unit uses the 3σ criterion to remove non-anomaly data, fills in one set of missing LED lighting LCA data using the KNN algorithm, and standardizes each data point using min-max normalization. The fusion computing unit uses an improved attention mechanism fusion algorithm to calculate the information entropy weights of each modality: BIM data 0.30, GIS data 0.18, environmental data 0.20, mechanical parameters 0.12, and LCA data 0.20, generating a standardized fusion dataset that is updated every 10 minutes.

[0047] 3. Dynamic carbon consumption prediction: Input the standardized fusion dataset and the renovation and construction schedule (exterior wall insulation construction 30 days, interior decoration renovation 40 days, equipment and pipeline renovation 20 days, completion 10 days) into the dynamic energy consumption-carbon emission coupling prediction module. A bidirectional LSTM network extracts the temporal correlation features between construction progress and energy consumption, outputting preliminary energy consumption predictions (external wall insulation construction energy consumption 3000kWh, interior decoration renovation energy consumption 4500kWh, equipment pipeline modification energy consumption 2500kWh); a GBRT model fits the nonlinear relationship between environmental parameters and carbon emissions, outputting preliminary carbon emission predictions (external wall insulation construction carbon emissions 35tCO2, interior decoration renovation carbon emissions 25tCO2, equipment pipeline modification carbon emissions 15tCO2); the coupled model obtains the final prediction result through adaptive weighted fusion. Compared with the preset threshold of the green building one-star standard, it is found that the carbon emission prediction value of the interior decoration renovation stage is 25tCO2, exceeding the threshold of 22tCO2, exceeding the standard by 13.6%, triggering an early warning and generating the warning message: "The carbon emission prediction value of the interior decoration renovation stage is 25tCO2, exceeding the standard by 13.6%. It is recommended to replace the floor decoration materials (replace the planned ordinary floor tiles with recycled plastic flooring) and select low-power construction equipment."

[0048] 4. Green Decision Generation: The intelligent material selection and supply chain collaboration module receives early warning information and fused datasets, calls upon the built-in green material database, and filters out recycled plastic flooring that meets interior decoration requirements (lifecycle carbon emissions reduced by 50% compared to ordinary floor tiles, wear resistance ≥ AC4 level, fire resistance B1 level). Multi-objective optimization is performed using the NSGA-Ⅲ algorithm, with input optimization objectives (lowest carbon emissions, optimal cost, shortest construction period) and constraints (flooring size 1220mm × 2440mm, delivery cycle ≤ 5 days, on-site storage capacity ≤ 100m³). 2 After 50 iterations, the algorithm outputs an optimal material combination list: recycled plastic flooring, water-based latex paint, light steel keel ceiling, and LED energy-saving lighting fixtures. The supply chain real-time status monitoring unit connects to the recycled plastic flooring supplier's ERP system to obtain the supplier's inventory (currently 1000m² of recycled plastic flooring). 2, meeting the project requirement of 800m 2 ), the location of the delivery vehicle (20 km away from the construction site), generate a delivery plan: "On [specific date] at 10:00 am, the delivery vehicle with license plate number Shanghai BXXX will deliver 800m 2 of recycled plastic flooring to the indoor warehouse of the construction site, avoiding direct sunlight", and push it to the supplier and the construction unit.

[0049] 5. Construction process control: After receiving the recycled plastic flooring, the construction unit collects material information through the RFID reader of the blockchain quality traceability module. After the verification of the supervision unit node, it is written into the blockchain. During the construction process, the IoT sensing devices collect the operating parameters of the construction machinery in real time (replaced with a low-power electric drill, rotation speed 2800 r / min, power 1.5 kW, energy consumption reduced by 25%) and environmental data, and upload them to the cross-modal data fusion engine to update the fusion dataset. The dynamic energy consumption-carbon emission coupling prediction module corrects the carbon emission prediction value in the indoor renovation stage to 21 tCO2, which is lower than the threshold. The management personnel monitor the construction progress through the visual decision terminal and find that the progress of the wall latex paint brushing process lags behind by 2 days, and issue a control instruction: "Add 2 construction workers and adopt a construction process combining roller coating and spraying to speed up the brushing progress". The quality inspection personnel collect data such as the wall brushing thickness (2 mm, meeting the design requirements) and the laying flatness of the recycled plastic flooring (deviation ≤ 2 mm), and upload them to the blockchain quality traceability module.

[0050] 6. Full-cycle traceability: After the project is completed, query the source of the recycled plastic flooring (supplier name, production batch), in-site inspection data, construction use parts (office floor, corridor floor) and green indicators (indoor VOC concentration 0.04 mg / m 3 after laying, meeting the standard, carbon emissions reduced by 12 tCO2 compared with ordinary floor tiles) through the blockchain quality traceability module; query the construction workers, construction time and quality inspection results of the wall latex paint brushing process. The traceability report provides strong support for the one-star green building review of the project. The project passes the review smoothly. After the renovation, the energy consumption of the office building is reduced by 30% compared with that before the renovation, and the indoor environmental quality is significantly improved.

[0051] In this embodiment, by adopting the system and method of the present invention, the actual carbon emission in the indoor renovation stage of the project is 20.5 tCO2, which is 18% lower than the predicted over-standard value and 19% lower than the traditional renovation method; the construction period is shortened by 8 days (7.7% shorter) compared with the plan; the material procurement cost is reduced by 6%; the green index compliance rate is 98%, achieving efficient management of the green renovation design and decoration project of existing buildings.

[0052] Example 3: Management of large public building decoration projects This example is applied to the decoration project of a large convention and exhibition center, with a building area of ​​50,000 square meters. 2 The complex includes functional areas such as exhibition halls, conference rooms, and office areas. Its design aims for a three-star green building standard, emphasizing low carbon emissions, environmental protection, energy conservation, and intelligent technology. The green building design and decoration engineering management system and method of this invention are used for full-cycle management. The specific implementation process is as follows: 1. System Deployment: The cross-modal data fusion engine adopts a cluster server architecture (2 Intel Xeon Gold 6248 processor servers, 64GB memory, 4TB hard drive), deploys multi-format BIM model parsing interfaces (supporting software formats such as Revit, Bentley, and SketchUp), and GIS map call interfaces (connecting to Google Maps API and domestic geographic information platforms). It also installs a large number of IoT sensing devices (50 temperature and humidity sensors, 20 wind speed sensors, 50 construction machinery operation parameter acquisition terminals, and 10 air quality sensors), achieving data transmission through a LoRa + 5G dual-mode gateway. The intelligent material selection and supply chain collaboration module connects to the international EPD database, the domestic green building materials database, and the ERP systems and logistics tracking systems of 5 large material suppliers. The blockchain quality traceability module deploys 6 consortium blockchain nodes (adding design unit nodes and third-party testing agency nodes), adopting the PBFT consensus mechanism to improve data consensus efficiency. The visualization decision-making terminal uses a 43-inch industrial touch screen and is installed in the construction site command center, construction unit, design unit, and supervision unit.

[0053] 2. Data Acquisition and Fusion: The cross-modal data fusion engine was activated to import the BIM model data of the convention center (including the structure of the large-span exhibition hall, complex decorative details, and intelligent equipment pipelines); GIS data of the project site was acquired (the convention center is located in a new urban area, 5km from the airport, with surrounding transportation hubs); IoT sensors collected real-time environmental data during construction (average temperature 26℃, humidity 60%, wind speed 3m / s), construction machinery operating parameters (large crane speed 20r / min, power 300kW, fuel consumption 50L / h; decoration robot speed 50r / min, power 20kW), and indoor air quality data (PM2.5 concentration 35μg / m³). 3 ); Obtain LCA data for materials and equipment such as high-performance energy-saving glass, aluminum honeycomb panels, eco-wood, and intelligent lighting systems through the LCA data interface (high-performance energy-saving glass lifecycle carbon emissions: 120 kg / m³). 2 The recyclability rate is 90%; the carbon emissions over the life cycle of aluminum honeycomb panels are 150 kg / m². 2(Recyclability rate 95%). The data preprocessing unit processes a large amount of collected data, removes 12 sets of abnormal mechanical parameter data, fills in 8 sets of missing material LCA data, and after standardization, calculates the weight of each modality data through an improved attention mechanism fusion algorithm: BIM data 0.25, GIS data 0.15, environmental data 0.20, mechanical parameters 0.18, LCA data 0.22, generating a standardized fusion dataset, which is updated every 5 minutes (due to the large scale of the project and the large amount of data, the update cycle is shortened).

[0054] 3. Dynamic Carbon Consumption Prediction: The fusion dataset and construction schedule (80 days for exhibition hall decoration, 30 days for conference room decoration, 20 days for office area decoration, and 20 days for completion) are input into the dynamic energy consumption-carbon emission coupled prediction module. A bidirectional LSTM network with 3 hidden layers and 128 neurons per layer is used for 200 iterations to extract temporal correlation features. A GBRT model is used to construct 100 regression trees to fit nonlinear relationships. The coupled model outputs the predicted energy consumption and carbon emission results for each stage. The predicted carbon emission value for the exhibition hall decoration stage is 350 tCO2, exceeding the three-star green building standard threshold of 320 tCO2 by 9.4%, triggering an early warning and generating the warning message: "The predicted carbon emission value for the exhibition hall decoration stage is 350 tCO2, exceeding the standard by 9.4%. It is recommended to replace the exhibition hall wall decoration materials (replace aluminum honeycomb panels with eco-wood composite panels), optimize the selection of the intelligent lighting system (use low-power LED light strips), and adjust the usage time of large machinery (avoid high-temperature periods to reduce energy consumption)."

[0055] 4. Green Decision Generation: The intelligent material selection and supply chain collaboration module selected eco-friendly wood composite panels (with a 30% reduction in lifecycle carbon emissions compared to aluminum honeycomb panels, strength ≥3MPa, and fire resistance rating A) and low-power LED light strips (power 5W / m, 20% lower energy consumption than ordinary LED light strips). Multi-objective optimization was performed using the NSGA-Ⅲ algorithm. Considering the large material demand and complex supply chain of large-scale projects, the optimization objective added "supply chain stability," with constraints including material supply capacity (daily supply of eco-friendly wood composite panels ≥500m³). 2 Delivery time (≤10 days), on-site storage capacity (≤1000m³) 2 After 80 iterations, the algorithm outputs an optimal material combination list. The real-time supply chain status monitoring unit connects with multiple suppliers to generate a batch delivery plan (the eco-wood composite board is delivered in 5 batches, each batch being 1000m²). 2 (with a 2-day interval), and coordinate logistics vehicle routes to avoid peak traffic periods.

[0056] 5. Construction Process Control: The construction unit receives materials according to a batch delivery plan. Material information and appearance quality are collected via RFID readers and high-definition cameras in the blockchain quality traceability module. A third-party testing agency conducts sampling tests on the materials, and data is written to the blockchain after passing the tests. During construction, IoT sensors collect real-time data on large machinery operating parameters (adjusting usage time to 6-11 AM and 4-8 PM to avoid the midday heat, reducing fuel consumption by 15%), environmental data, and indoor air quality data. This data is uploaded to the cross-modal data fusion engine to update the fused dataset. The dynamic energy consumption-carbon emission coupling prediction module corrects the carbon emission prediction value for the exhibition hall decoration construction phase to 315 tCO2, below the threshold. Management personnel monitor the progress and quality of multiple construction areas through a visual decision-making terminal and coordinate construction resources in each area through intelligent scheduling to avoid resource conflicts. Quality inspection personnel use professional testing equipment to collect data on exhibition hall floor flatness, wall verticality, and intelligent lighting system illuminance, which is then uploaded to the blockchain quality traceability module.

[0057] 6. Full-cycle traceability: After project completion, the blockchain quality traceability module enables full-cycle traceability of large-scale projects, allowing for multi-dimensional queries based on functional areas (exhibition hall, meeting rooms, office areas), material types, construction units, and more. It allows querying of supply information, test reports, application locations (exhibition hall walls 1-5), and green indicators (carbon emissions reduced by 105tCO2 compared to aluminum honeycomb panels, indoor formaldehyde concentration 0.02mg / m³) for each batch of eco-wood composite panels. 3 The system's installation and commissioning data, as well as energy consumption monitoring data, were retrieved (the average illuminance in the exhibition hall was 500 lx, and energy consumption was reduced by 25% compared to ordinary lighting systems). The traceability report provided a comprehensive and accurate basis for the project's three-star green building assessment, and the project successfully passed the assessment, becoming a local green building demonstration project.

[0058] In this embodiment, by adopting the system and method of the present invention, the actual carbon emissions during the exhibition hall decoration construction phase were 310tCO2, which is 11.4% lower than the predicted excess value and 20% lower than the traditional construction method; the construction period was shortened by 20 days (9.5%) compared to the plan; material procurement costs were reduced by 8%; and the green indicator compliance rate was 99%, realizing refined and green management of large-scale public building decoration projects.

[0059] Comparison Example To verify the beneficial effects of the present invention, the following three sets of comparative examples were set up, using existing technology to carry out the same type of green building design and decoration engineering management, with other conditions being consistent with the corresponding embodiments.

[0060] Compare with Example 1: New green residential projects using traditional management systems (single schedule management based on BIM) The project only simulated the progress using the BIM model, without multi-source data fusion or dynamic carbon consumption prediction. Material selection relied on the experience-based judgment of management personnel, without referencing the LCA (Life Cycle Assessment) data of green materials, and instead selected non-green materials such as ordinary ceramic tiles and solvent-based coatings based solely on price and conventional performance. During construction, only the construction progress was monitored, without real-time control over energy consumption and carbon emissions; post-construction accounting was only conducted after project completion using construction records and material consumption data. Quality traceability relied on paper records combined with a centralized database, with material arrival inspection data and construction process quality data being entered unilaterally by the construction unit, making data tampering and missing information prone to occurrence.

[0061] Compare with Example 2: Existing building green renovation projects use ordinary renovation management systems (without reverse BIM integration and dynamic control functions). The existing conventional building renovation management system was used, but no reverse BIM model of the existing building was built, making it impossible to integrate the original building structure data with the renovation construction data; the material selection was not connected to the LCA database, and only common common insulation materials and floor tiles were selected, without considering the carbon emissions of the materials' life cycle; energy consumption data was recorded manually on a regular basis during construction, without real-time monitoring and early warning functions, and carbon emission control relied on post-construction rectification; quality traceability only covered the construction stage, not the material procurement and design stages, and the data was stored in a centralized database, which posed a risk of data tampering.

[0062] Compare with Example 3: Large-scale public building decoration projects use traditional large-scale project management systems (lacking multi-source data integration and blockchain traceability functions). The existing traditional management system for large public building projects can only process BIM or GIS data separately, and cannot effectively integrate multi-source heterogeneous data; energy consumption and carbon emission predictions use a single linear model, which does not consider the coupled effects of environmental parameters, construction progress and material performance, resulting in large prediction errors; material selection does not undergo multi-objective optimization, but only prioritizes cost factors, and uses materials such as ordinary aluminum honeycomb panels and traditional lighting systems; supply chain management uses manual interaction and lacks real-time status monitoring functions, which easily leads to material supply disruptions; quality traceability only sets up two nodes, the construction unit and the contractor, without the participation of design units and third-party testing agencies, resulting in poor data consensus, and the use of traditional database storage cannot guarantee data security and immutability.

[0063] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of protection claimed by the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A green building design and decoration engineering management system, characterized in that, include: The cross-modal data fusion engine collects BIM data from building design, site GIS data, environmental IoT sensor data, and LCA lifecycle data from green materials. It employs an improved attention-based fusion algorithm to clean, align, and fuse multi-source heterogeneous data, outputting a standardized fused dataset. The dynamic energy consumption-carbon emission coupled prediction module communicates with the cross-modal data fusion engine. Based on the fused dataset, it constructs a bidirectional LSTM-GBRT coupled model, inputting construction schedules, environmental parameters, and material usage plans, and outputs real-time predictions of energy consumption and carbon emissions at each stage of the project. The intelligent material selection and supply chain collaboration module connects with both the cross-modal data fusion engine and the dynamic energy consumption-carbon emission coupled prediction module. The system includes an energy consumption-carbon emission coupled prediction module, a built-in green materials database and a real-time supply chain status monitoring unit. Based on the prediction results and engineering design requirements, it uses a multi-objective optimization algorithm to select the optimal combination of green materials and generate a supply chain distribution plan. A blockchain quality traceability module, connected to the intelligent material selection and supply chain collaboration module and the construction monitoring unit, employs a consortium blockchain architecture to write material arrival inspection data, construction process quality data, and green indicator compliance data into blockchain nodes, achieving tamper-proof data traceability. A visual decision-making terminal communicates with each of the above modules to display integrated data, prediction results, material selection schemes, and traceability information, and supports management personnel in issuing control instructions.

2. The green building design and decoration engineering management system according to claim 1, characterized in that: The cross-modal data fusion engine includes a data acquisition unit, a data preprocessing unit, and a fusion computing unit. The data acquisition unit is equipped with a BIM model parsing interface, a GIS map calling interface, an IoT sensor gateway, and an LCA data docking interface, supporting the simultaneous acquisition of three-dimensional geometric data, geospatial data, real-time environmental data, and material life cycle data.

3. The green building design and decoration engineering management system according to claim 1, characterized in that: The bidirectional LSTM-GBRT coupling model in the dynamic energy consumption-carbon emission coupled prediction module uses a bidirectional LSTM network to extract the correlation features between construction progress and energy consumption in the time dimension, and fits the nonlinear relationship between environmental parameters and carbon emissions through the GBRT model. The output results of the two are then weighted and fused to obtain the final prediction value.

4. The green building design and decoration engineering management system according to claim 1, characterized in that: The multi-objective optimization algorithm of the intelligent material selection and supply chain collaboration module aims to minimize carbon emissions, optimize costs, and shorten the construction period. The constraints include material performance compliance, supply chain delivery capabilities, and on-site storage capacity.

5. The green building design and decoration engineering management system according to claim 1, characterized in that: The blockchain quality traceability module includes a data acquisition terminal, a consortium blockchain node server, and a traceability query unit; the consortium blockchain nodes include construction unit nodes, construction unit nodes, supervision unit nodes, and material supplier nodes, and each node uses the PBFT consensus mechanism to achieve data consensus verification.

6. A method for managing green building design and decoration projects, wherein the method is implemented using the green building design and decoration project management system as described in any one of claims 1-5, characterized in that, Includes the following steps: S1. Data Acquisition and Fusion: The cross-modal data fusion engine collects architectural design BIM data, site GIS data, environmental IoT sensor data, and green material LCA data. After preprocessing, an improved attention mechanism fusion algorithm is used to generate a standardized fusion dataset. S2. Dynamic carbon consumption prediction: Input the fused dataset and construction schedule into the bidirectional LSTM-GBRT coupled model to predict the energy consumption and carbon emissions at each stage of the project in real time. If the predicted value exceeds the preset green threshold, an early warning will be triggered. S3, Green Decision Generation: The intelligent material selection and supply chain collaboration module uses a multi-objective optimization algorithm to select green material combinations and generate material procurement and distribution plans based on prediction results and design requirements. S4. Construction process control: During the construction process, process quality data and green indicator compliance data are collected in real time and uploaded to the blockchain quality traceability module. Managers monitor the construction status through a visual decision terminal and adjust the construction plan according to the early warning information. S5. Full-cycle traceability: After the project is completed, the blockchain quality traceability module enables full-cycle traceability and query of material sources, construction procedures, and compliance with green indicators.

7. The green building design and decoration engineering management method according to claim 6, characterized in that: The improved attention mechanism fusion algorithm in step S1 calculates the information entropy weights of each modality data and performs weighted fusion of data from different sources. The weight update cycle is consistent with the IoT data collection cycle.

8. A method for managing green building design and decoration projects according to claim 6, characterized in that: If the predicted value exceeds the preset threshold in step S2, the triggered warning information includes the type of exceedance, the extent of exceedance, and the suggested adjustment plan. The suggested adjustment plan includes material replacement suggestions and construction process optimization suggestions.

9. A method for managing green building design and decoration projects according to claim 6, characterized in that: The construction process control in step S4 also includes real-time monitoring of the energy consumption of construction machinery. This involves collecting machinery operating parameters through IoT sensor nodes, calculating the deviation between actual and predicted energy consumption, and adjusting the machinery usage plan in a timely manner.

10. A method for managing green building design and decoration projects according to claim 6, characterized in that: The full-cycle traceability query in step S5 supports multi-dimensional queries by engineering stage, material type, and quality inspection items, and the query results generate a traceability report with timestamps.

Images