Construction whole-process carbon emission metering method based on multi-source data fusion and dynamic feedback correction

Abstract

The invention discloses a construction whole-process carbon emission metering method based on multi-source data fusion and dynamic feedback correction. The method comprises the steps that multi-source data such as construction equipment operation, building material logistics and environmental parameters are collected through Internet of Things equipment and preprocessed; using the improved neural network model to classify the associated data, and outputting the comprehensive energy consumption and carbon emission potential of the construction link; a metering model covering transportation, equipment and personnel activities is constructed, and link calculation is achieved in combination with dynamic parameters; and through real-time CO2 monitoring and comparison with a predicted value, analyzing a positioning deviation and correcting the model based on a fault tree. Through multi-source data fusion and dynamic feedback correction architecture, total element coverage and real-time self-adaptive correction of carbon emission in each construction stage are realized, the metering precision is improved, a construction unit is supported to quickly identify a high-carbon link and optimize the process, and the method is high in compatibility, can adapt to different construction projects and has remarkable engineering application value.

Description

Technical Field

[0001] The present invention relates to the technical field of building construction technology, and in particular to a method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction. Background Art

[0002] As global climate change becomes increasingly serious, the construction industry, as a key area of ​​carbon emissions, has made accurate measurement of carbon emissions during the construction phase a key link. The construction process involves multiple links such as building materials production and transportation, equipment operation, and personnel activities. The factors affecting carbon emissions are complex, and traditional measurement methods are difficult to meet the full-process, high-precision monitoring needs.

[0003] Currently, the following methods are mainly used to measure construction carbon emissions:

[0004] Calculations rely on building materials lists combined with industry average carbon emission factors, or only collect equipment energy consumption data for statistics. Such methods ignore key factors such as environmental parameters and construction process differences, resulting in measurement results that deviate from actual results by more than 20%.

[0005] Carbon emissions are regularly summarized by manually entering data such as construction logs and energy bills. However, this model has problems such as data lag, large manual errors, and inability to respond to dynamic changes in construction in real time, making it difficult to support low-carbon construction process management.

[0006] The carbon emissions of each construction link are calculated independently and then accumulated without considering the correlation between multi-source data. In addition, the model parameters are fixed, and the measurement accuracy is significantly reduced in complex construction scenarios.

[0007] Therefore, the prior art has the following defects:

[0008] First, there is a lack of integration of construction drawings, real-time data from IoT devices, and environmental monitoring data, making it impossible to fully capture the factors affecting carbon emissions, resulting in a single data dimension.

[0009] Secondly, a differentiated measurement model for construction stages and links has not been established, and there is no effective correction mechanism for the deviation between actual carbon emissions and predicted values, resulting in insufficient dynamic adaptability;

[0010] In addition, traditional methods are susceptible to noise interference when pursuing real-time performance, and high-precision measurement relies on a large amount of historical data training, which makes it difficult to take into account the efficiency of engineering implementation. Therefore, there is a contradiction between real-time performance and accuracy.

[0011] Therefore, how to provide a carbon emission measurement method for the entire construction process based on multi-source data fusion and dynamic feedback correction is an urgent problem that technicians in this field need to solve. Summary of the Invention

[0012] One purpose of the present invention is to propose a method for measuring carbon emissions throughout the construction process based on multi-source data fusion and dynamic feedback correction. The present invention can integrate multi-source data such as construction drawings, equipment operation, and environmental parameters, and combine intelligent fusion algorithms with dynamic feedback mechanisms to achieve accurate measurement and real-time correction of carbon emissions at all stages and links of construction, thereby solving the problems of traditional methods such as single data, poor dynamic adaptability, and insufficient measurement accuracy, and providing reliable technical support for carbon emission management in construction and carbon assessment throughout the entire life cycle.

[0013] The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to an embodiment of the present invention includes the following steps:

[0014] S1. Multi-source data collection and preprocessing: IoT devices collect construction drawings, building material purchase lists, machinery and equipment operation data, construction personnel attendance data, environmental monitoring data, energy consumption data, and carbon emission monitoring equipment data in real time, and then clean, remove noise, and convert the collected data into different formats.

[0015] S2. Multi-source data fusion: Classify the pre-processed data by construction stage, construction link, and data type, and establish correlations. Use fusion algorithms based on neural networks, fuzzy logic, or Bayesian networks to build a data fusion model and output comprehensive data for each construction stage and link.

[0016] S3. Carbon emission measurement model construction: Clarify the carbon emission measurement boundaries of the entire construction process, establish corresponding carbon emission calculation models for the production and transportation of building materials, operation of construction equipment, and activities of construction personnel, and determine model parameters through the fusion of multi-source data;

[0017] S4. Dynamic feedback correction: Use carbon emission monitoring equipment to collect actual carbon emission data in real time, compare it with the data predicted by the measurement model to calculate the deviation, analyze the cause of the deviation and dynamically correct the measurement model. Set the feedback correction cycle according to the project scale;

[0018] S5. Result output and application: Generate carbon emission reports for each stage and link of the entire construction process based on the revised measurement model, which will be used for carbon emission management of construction projects and carbon emission assessment of the entire life cycle of buildings.

[0019] Furthermore, the environmental monitoring data in step S1 includes wind speed, temperature, and humidity data, and during data preprocessing, the environmental monitoring data is denoised using a Kalman filter algorithm to eliminate the impact of random environmental interference on the data.

[0020] Furthermore, the energy consumption data in step S1 includes electricity, fuel oil, and gas consumption data. When collecting electricity consumption data, a load decomposition algorithm is used to decompose the total electricity consumption into each electrical device to achieve accurate measurement.

[0021] Furthermore, the construction stages in step S2 include foundation construction, main construction, and decoration construction stages. When fusing multi-source data, different data fusion weights are set for different construction stages. The weights of mechanical equipment operation data and building materials transportation data are increased in the foundation construction stage, and the weights of energy consumption data and construction personnel activity data are increased in the decoration construction stage.

[0022] Furthermore, the construction process in step S2 includes earth excavation, concrete pouring, and equipment installation. In the construction of the carbon emission measurement model, the carbon emission calculation model of the earth excavation process introduces a soil disturbance coefficient, which is dynamically adjusted according to the soil type and excavation depth; the model of the concrete pouring process takes into account the carbon emissions during the concrete curing process, and calculates additional carbon emissions based on the curing method and curing time.

[0023] Furthermore, in the dynamic feedback correction of step S4, when the deviation between the actual carbon emission data and the predicted data exceeds the set threshold, the deviation analysis and model correction are triggered. During the deviation analysis, the fault tree analysis method is used to check the cause of the deviation layer by layer to determine the specific factors causing the deviation.

[0024] Furthermore, the threshold is set to 10%, and when the deviation exceeds the threshold in two consecutive feedback cycles, the reconstruction program of the measurement model is automatically started to rebuild the carbon emission measurement model.

[0025] Furthermore, the feedback correction cycle in step S4 is daily, weekly or after each stage of construction. In the feedback correction after each stage of construction, the measurement model parameters of the next stage are pre-adjusted in combination with the construction experience data of this stage to improve the prediction accuracy.

[0026] Furthermore, in step S2, transfer learning technology is used to migrate historical data of similar construction projects to the data fusion model of the current project, thereby accelerating model training and improving fusion accuracy.

[0027] Furthermore, in step S5, the carbon emission report is displayed in the form of a three-dimensional dynamic chart using visualization technology, and the temporal and spatial variation trends of carbon emissions in each construction stage and link can be intuitively viewed through the chart.

[0028] The beneficial effects of the present invention are:

[0029] 1. The present invention uses IoT devices to collect multi-source data such as construction drawings, equipment operation, building materials logistics, and environmental parameters in real time, solving the problem of information loss caused by traditional methods relying on a single data source, and achieving full coverage of carbon emissions throughout the construction process. It also uses an improved neural network fusion model to perform layered feature extraction on multi-source data, effectively filtering out noise and strengthening key carbon emission influencing factors, thereby improving measurement accuracy. Secondly, it automatically adjusts the data fusion weight for different stages such as foundation construction, main construction, and decoration construction to adapt to changes in the construction process and avoid deviations caused by one-size-fits-all measurement.

[0030] 2. The present invention uses real-time comparison between high-precision CO2 monitors and metering models, combined with the fault tree analysis method to locate the cause of deviation layer by layer, ensuring that the metering results are updated synchronously with the actual construction status, and solving the defects of offline calculation and delayed feedback of traditional methods. A dual correction strategy of parameter adjustment and model reconstruction and automatic update of local parameters such as equipment energy consumption coefficient in case of slight deviation are set. When the continuous deviation exceeds the threshold, model reconstruction is initiated, and the neural network is retrained based on real-time data to improve the adaptability of the model in scenarios such as construction plan changes and sudden environmental interference. The feedback cycle is day / week / stage. After the completion of the stage construction, the historical experience data is used to pre-adjust the model parameters of the next stage, forming a closed-loop management of monitoring-analysis-correction-prediction to continuously optimize the metering accuracy.

[0031] 3. The present invention introduces the Kalman filter algorithm to process environmental data noise, adopts the load decomposition algorithm to accurately locate the power consumption of each device, and dynamically corrects the carbon emissions of earth excavation in combination with the soil disturbance coefficient. It solves the complex interference problems of construction scenarios through the cross-disciplinary technology, and uses the historical data migration of similar projects to initialize the fusion model parameters, so as to reduce the model training time of new projects and avoid repeated modeling costs. It is particularly suitable for the rapid implementation of construction projects such as residential and commercial complexes with a high degree of standardization. It also intuitively displays the distribution and change trends of carbon emissions in each stage through three-dimensional dynamic charts, supports construction units to quickly locate high-carbon links, provides data support for low-carbon process optimization, and helps to achieve carbon emission reduction during the construction stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0033] Figure 1 This is a flow chart of the method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction proposed by the present invention;

[0034] Figure 2This is a diagram of the multi-source data fusion model architecture of the method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction proposed in the present invention;

[0035] Figure 3 This is the dynamic feedback correction logic diagram of the construction process carbon emission measurement method based on multi-source data fusion and dynamic feedback correction proposed by the present invention. DETAILED DESCRIPTION

[0036] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0037] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0038] It should be noted that, in the description of the present invention, the terms "construction stage" and "construction link" are based on the conventional division of the construction process and are only for the convenience of description, and are not limitations on the technical solution of the present invention.

[0039] In addition, it should be noted that in the description of the present invention, technical terms such as "multi-source data fusion" and "dynamic feedback correction" should be understood in accordance with the general technical meanings in this field. During specific implementation, appropriate hardware equipment and software algorithms can be selected according to the characteristics of the project.

[0040] See also Figure 1-3 As shown, the method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to an embodiment of the present invention includes the following steps:

[0041] S1: Multi-source data collection and preprocessing

[0042] Data collection equipment is deployed on the construction site to deploy an IoT sensor network, including:

[0043] Building materials management module: GPS locators and weight sensors are installed on building materials transportation vehicles to collect building materials purchase lists and transportation route data in real time; RFID readers are installed at the warehouse entrance to automatically identify building materials entry information.

[0044] Equipment monitoring module: Install an OBD-II on-board diagnostic system on construction machinery to collect real-time equipment operating data, including speed, operating hours, and fuel consumption. Install a smart meter in the distribution box and use a load decomposition algorithm to break down total power consumption into individual electrical equipment, such as lighting systems, construction elevators, and welding machines, to achieve accurate measurement of power consumption.

[0045] Environmental monitoring module: A micro-meteorological station is deployed at the construction site to collect wind speed, temperature, and humidity data in real time, and upload the monitoring values ​​every 5 minutes.

[0046] Personnel management module: The attendance data of construction personnel is collected through the facial recognition attendance system, and the personnel activity areas and working hours are obtained in combination with the positioning bracelet.

[0047] Carbon emission monitoring module: Install infrared spectrometer gas analyzers at fixed pollution sources, such as generator exhaust outlets, to monitor the concentrations of greenhouse gases such as CO2 and CH4 in real time; install portable carbon emission monitors on mobile machinery to analyze carbon emission data through exhaust gas sampling.

[0048] Data preprocessing process

[0049] Data cleaning: Use the Python Pandas library to remove duplicate data and verify business rules, such as the number of building materials is non-negative and the equipment speed is within a reasonable range, to eliminate outliers.

[0050] Denoising: The Kalman filter algorithm is used for environmental monitoring data. The state equation is set as x(k) = x(k-1) + w(k), and the measurement equation is z(k) = x(k) + v(k). Where w(k) and v(k) are Gaussian white noise. Recursive filtering is performed with a window of 10 time steps to eliminate random environmental interference.

[0051] Format conversion: All data is uniformly converted into JSON format, including metadata such as timestamp, data type, construction stage, construction link, equipment number / personnel ID, and stored in a distributed database.

[0052] S2. Multi-source data fusion

[0053] Data classification and association

[0054] Construction phase division: The project is divided into three phases: foundation construction, main construction, and decoration construction. An independent data fusion space is established for each phase. Foundation construction includes pile foundation engineering, earth excavation, and basement construction. Main construction includes reinforced concrete structure and steel structure installation. Decoration construction includes wall masonry, decorative decoration, and equipment installation.

[0055] Data association model: Construct a three-dimensional data association table of stage-link-type. For example, in the earth excavation stage of the foundation construction phase, the associated data types include mechanical equipment operation data, building materials transportation data, and environmental data.

[0056] Dynamic weight fusion algorithm

[0057] Weight configuration strategy:

[0058] Foundation construction stage: Mechanical equipment operation data (weight 0.4) > Building materials transportation data (0.3) > Energy consumption data (0.2) > Personnel data (0.1), because the mechanical operation intensity is high and the transportation of building materials is frequent during this stage.

[0059] During the main construction phase: energy consumption data (0.35) > machinery and equipment data (0.3) > building materials data (0.25) > environmental data (0.1), focusing on energy-consuming links such as concrete pouring and steel bar processing.

[0060] During the decoration and construction phase: energy consumption data (0.4) > personnel activity data (0.3) > equipment data (0.2) > building materials data (0.1), focusing on indirect emissions from indoor decoration electricity consumption and personnel operations.

[0061] Fusion model construction: An improved BP neural network is used, with the number of input layer nodes dynamically adjusted according to the data type. The hidden layer is set to 2 layers, and the activation function is ReLU. An independent model is trained for each construction stage. At the same time, transfer learning technology is introduced to migrate historical fusion model parameters of similar projects, such as buildings of the same type and similar scale, to the current project, reducing the initial training sample size by 50% and increasing the training speed by 30%.

[0062] S3. Construction of carbon emission measurement model

[0063] Measurement boundary definition

[0064] Spatial boundary: covers the red line of the construction site, including temporary facilities, construction machinery, and personnel activity areas; extends to the production and transportation of building materials, from the building materials factory to the construction site.

[0065] Time boundary: From the start of foundation construction to the end of decoration construction, the time units are divided into days / weeks / construction stages.

[0066] Link model establishment

[0067] Taking earth excavation as an example: basic carbon emissions are calculated by multiplying the fuel consumption of equipment such as excavators by the diesel carbon emission factor; among them, the equipment fuel consumption can be obtained from the data collected by the fuel consumption sensor, and the diesel carbon emission factor adopts the industry standard value of 2.68kgCO2 / L.

[0068] On this basis, the soil disturbance coefficient is introduced. This coefficient is dynamically adjusted according to the soil type and excavation depth. The soil disturbance coefficient formula is K = α × D + β × S, where D is the excavation depth, S is the soil hardness coefficient, α = 0.2, and β = 0.05.

[0069] Taking concrete curing as an example: the additional carbon emissions generated by watering curing are calculated by multiplying the curing water consumption by the water treatment carbon emission factor. The curing water consumption can be obtained based on actual records, and the water treatment carbon emission factor is 0.91kgCO2 / m 3 .

[0070] S4, dynamic feedback correction

[0071] Real-time deviation calculation

[0072] Create a real-time data comparison dashboard, synchronize monitoring data with model predictions every hour, and calculate the deviation rate:

[0073] When the deviation rate exceeds the 10% threshold for three consecutive times, once every 10 minutes, a yellow warning is triggered; if the deviation exceeds the threshold for two consecutive feedback cycles, a red warning is triggered and model reconstruction is initiated.

[0074] Deviation analysis and correction

[0075] Fault Tree Analysis (FTA): Starting from the top event (deviation exceeding threshold), the system decomposes the event layer by layer into intermediate events such as the data layer (acquisition error, fusion error), the model layer (outdated parameters, missing boundaries), and the environment layer (sudden weather changes, process changes), ultimately locating the bottom event, such as a sensor failure in a certain equipment or an error in the calculation of maintenance time.

[0076] Correction strategy:

[0077] Data layer issues: Automatically mark abnormal data points, trigger a manual verification process, and re-enter the model after correction.

[0078] Model-level issues: Adjust model parameters individually for links where deviations exceed the standard, such as increasing the emission factor of a certain type of machinery. If the cumulative number of corrections exceeds 5 times / stage, initiate partial reconstruction of the model.

[0079] End-of-stage correction: After the foundation construction phase is completed, the actual data of this phase, such as the total amount of earth excavation and machine efficiency, are used to pre-adjust the model parameters of the main construction phase. For example, the energy consumption weight of the concrete pouring stage is increased by 10% to improve the prediction accuracy of the next phase.

[0080] S5. Output and apply results to generate multi-dimensional carbon emission reports

[0081] Report content: Specifically, the report includes:

[0082] Basic data: total carbon emissions and intensity (kgCO2 / ㎡) at each stage / link;

[0083] Composition analysis: the proportion of dimensions such as equipment operation, building materials transportation, and personnel activities;

[0084] Trend curve: weekly trend of carbon emissions during the construction period, marking key deviation points and corrective measures.

[0085] Visualization, specifically, using bird's-eye view, timeline view, and comparison view to form a three-dimensional dynamic chart:

[0086] The bird's-eye view displays a real-time carbon emissions heat map for each area of ​​the construction site, with red areas indicating high carbon emissions. The timeline view displays a bar chart of carbon emissions for each construction stage, allowing drill-down to view data for specific links. The comparison view displays a dynamic comparison curve between actual and budgeted carbon emissions, with an automatic red alert when the deviation exceeds 5%.

[0087] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0088] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction, characterized by: The following steps are involved: S1. Multi-source data collection and preprocessing: IoT devices collect construction drawings, building material purchase lists, machinery and equipment operation data, construction personnel attendance data, environmental monitoring data, energy consumption data, and carbon emission monitoring equipment data in real time, and then clean, remove noise, and convert the collected data into different formats. S2. Multi-source data fusion: Classify the pre-processed data by construction stage, construction link, and data type, and establish correlations. Use fusion algorithms based on neural networks, fuzzy logic, or Bayesian networks to build a data fusion model and output comprehensive data for each construction stage and link. S3. Carbon emission measurement model construction: Clarify the carbon emission measurement boundaries of the entire construction process, establish corresponding carbon emission calculation models for the production and transportation of building materials, operation of construction equipment, and activities of construction personnel, and determine model parameters through the fusion of multi-source data; S4. Dynamic feedback correction: Use carbon emission monitoring equipment to collect actual carbon emission data in real time, compare it with the data predicted by the measurement model to calculate the deviation, analyze the cause of the deviation and dynamically correct the measurement model. Set the feedback correction cycle according to the project scale; S5. Result output and application: Generate carbon emission reports for each stage and link of the entire construction process based on the revised measurement model, which will be used for carbon emission management of construction projects and carbon emission assessment of the entire life cycle of buildings.

2. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: The environmental monitoring data in step S1 includes wind speed, temperature, and humidity data, and during data preprocessing, the environmental monitoring data is denoised using a Kalman filter algorithm to eliminate the impact of random environmental interference on the data.

3. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: The energy consumption data in step S1 includes electricity, fuel oil and gas consumption data. When collecting electricity consumption data, a load decomposition algorithm is used to decompose the total electricity consumption into each electrical device to achieve accurate measurement.

4. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: The construction stages in step S2 include foundation construction, main construction, and decoration construction stages. When fusing multi-source data, different data fusion weights are set for different construction stages. The weights of mechanical equipment operation data and building material transportation data are increased in the foundation construction stage, and the weights of energy consumption data and construction personnel activity data are increased in the decoration construction stage.

5. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: The construction process in step S2 includes earth excavation, concrete pouring, and equipment installation. In the construction of the carbon emission measurement model, the carbon emission calculation model of the earth excavation process introduces a soil disturbance coefficient, which is dynamically adjusted according to the soil type and excavation depth; the model of the concrete pouring process takes into account the carbon emissions during the concrete curing process and calculates additional carbon emissions based on the curing method and curing time.

6. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: In the dynamic feedback correction of step S4, when the deviation between the actual carbon emission data and the predicted data exceeds the set threshold, the deviation analysis and model correction are triggered. During the deviation analysis, the fault tree analysis method is used to check the cause of the deviation layer by layer to determine the specific factors causing the deviation.

7. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 6 is characterized in that: The threshold is set to 10%, and when the deviation exceeds the threshold in two consecutive feedback cycles, the reconstruction program of the measurement model is automatically started to rebuild the carbon emission measurement model.

8. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: The feedback correction cycle in step S4 is daily, weekly or after each stage of construction. In the feedback correction after each stage of construction, the measurement model parameters of the next stage are pre-adjusted in combination with the construction experience data of the current stage to improve the prediction accuracy.

9. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: In step S2, transfer learning technology is used to migrate historical data of similar construction projects to the data fusion model of the current project, thereby accelerating model training and improving fusion accuracy.

10. The method for measuring carbon emissions during the entire construction process based on multi-source data fusion and dynamic feedback correction according to claim 1 is characterized in that: In step S5, the carbon emission report is displayed in the form of a three-dimensional dynamic chart using visualization technology, and the temporal and spatial variation trends of carbon emissions in each construction stage and link can be intuitively viewed through the chart.

Images