A low-carbon construction scheduling optimization method, device, equipment and storage medium

Abstract

The application discloses a low-carbon construction scheduling optimization method, comprising the following steps: collecting multi-source data through a multi-source heterogeneous sensor network, and constructing a standardized time series dataset; identifying the running state of construction equipment based on the time series dataset, and generating energy consumption data of each construction equipment in each running state; obtaining corresponding dynamic carbon emission factors according to the energy type of the construction equipment, combining the dynamic carbon emission factors and the energy consumption data, and calculating the device-level carbon emission power; based on a preset mapping relationship, mapping the device-level carbon emission power to the process level and the component level, and outputting multi-scale carbon emission data; based on the time series dataset and the multi-scale carbon emission data, constructing a Bayesian network, and identifying key factors affecting carbon emission; constructing a multi-objective optimization model, and generating a low-carbon construction scheduling strategy in combination with the key factors. The application can realize full-link tracing and scheduling optimization of construction carbon emission, and provides reliable technical support for green and low-carbon development of building construction.

Description

Technical Field

[0001] This invention relates to the field of low-carbon construction scheduling technology, and in particular to a low-carbon construction scheduling optimization method, apparatus, equipment and storage medium. Background Technology

[0002] With the advancement of the "dual carbon" strategy, the construction phase has become the stage with the least real-time monitoring capability, the largest error, and the highest management difficulty in the carbon emissions of a building throughout its entire life cycle.

[0003] Existing methods generally rely on construction lists, shift records, or process quantities for emission estimation, which cannot reflect real-time energy consumption behavior at construction sites, capture the large amount of hidden waste caused by equipment standby, idling, overloading, and repeated start-stop, and cannot analyze the causes of emission anomalies or optimize on-site scheduling. At the same time, the diverse types of equipment, uncertain usage, and large fluctuations in energy supply at construction sites make the traditional "fixed emission factor + process quantification" calculation method insufficient in accuracy and have a long update cycle, making it difficult to meet the needs of high-frequency carbon audits and green construction management. Summary of the Invention

[0004] This invention provides a low-carbon construction scheduling optimization method that can solve the problems of extensive carbon emission control and lack of data support in traditional construction scheduling optimization. It realizes full-link traceability and scheduling optimization of construction carbon emissions, and provides reliable technical support for the green and low-carbon development of building construction.

[0005] In a first aspect, embodiments of the present invention provide a low-carbon construction scheduling optimization method, comprising: Multi-source data is collected by a multi-source heterogeneous sensor network deployed at the construction site, the multi-source data is preprocessed, and a standardized time-series dataset is constructed. Based on the time-series dataset, the operating status of construction equipment is identified, and energy consumption data of each piece of construction equipment in each operating status is generated. The dynamic carbon emission factor is obtained according to the energy type of the construction equipment, and the equipment-level carbon emission power is calculated by combining the dynamic carbon emission factor and the energy consumption data. Based on a preset mapping relationship, the equipment-level carbon emission power is mapped to the process level and component level, and multi-scale carbon emission data is output. A Bayesian network is constructed based on the time-series dataset and the multi-scale carbon emission data, and the Bayesian network is used to identify key factors affecting carbon emissions. A multi-objective optimization model is constructed, and a low-carbon construction scheduling strategy is generated by combining the aforementioned key factors.

[0006] Furthermore, the preprocessing of the multi-source data includes: The collected multi-source data is resampled to a uniform time interval; Anomaly detection is performed on resampled multi-source data to identify and mark outliers in the data; For detected abnormal and missing data, corresponding repair methods are used to repair them.

[0007] Furthermore, the step of identifying the operating status of construction equipment based on the time-series dataset and generating energy consumption data for each piece of construction equipment in each operating status includes: Based on the time-series dataset, equipment operation feature parameters are extracted to identify the operating status of the construction equipment. The operating status includes at least the shutdown status, standby status, light load operating status, and heavy load operating status. The time-series dataset is segmented based on the transition boundaries of the operating state to form energy consumption data segments that correspond one-to-one with the operating state of the equipment. The cumulative energy consumption value within each energy consumption data segment is calculated to obtain the energy consumption data of each construction equipment in each operating state. The energy consumption data includes equipment identification, operating state, time period information, and cumulative energy consumption value.

[0008] Furthermore, the step of obtaining the corresponding dynamic carbon emission factor based on the energy type of the construction equipment, and calculating the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data, includes: Extract the energy type identifier, time period information, and energy consumption value of the construction equipment from the energy consumption data; Based on the energy type identifier and the time period information, the dynamic carbon emission factor of the corresponding energy type in the corresponding time period is queried from the pre-built carbon emission factor time series database; wherein, the carbon emission factor time series database is used to store the carbon emission factor values ​​of a specific energy type in a specific time period; Based on the energy consumption value and the dynamic carbon emission factor, the equipment-level carbon emission power is calculated.

[0009] Furthermore, based on a preset mapping relationship, the equipment-level carbon emission power is mapped to the process-level and component-level, outputting multi-scale carbon emission data, including: Based on the preset equipment-process mapping relationship, the carbon emission power of all related equipment in each construction process is summed by time integral to obtain process-level carbon emission data. Based on the preset process-component mapping relationship, the process-level carbon emission data is allocated to each associated component according to the weight ratio of the engineering quantity, thereby generating component-level carbon emission data; By integrating carbon emission data at the equipment, process, and component levels, multi-scale carbon emission data is formed, which includes multi-dimensional identifiers, time period information, and carbon emission values.

[0010] Furthermore, the Bayesian network constructed based on the time-series dataset and the multi-scale carbon emission data is used to identify key factors affecting carbon emissions, including: Factors related to carbon emissions are selected from the time-series dataset as observation nodes of the Bayesian network, and process-level carbon emissions are used as target nodes of the Bayesian network. The Bayesian network is constructed based on the causal relationships between the nodes. Based on the constructed Bayesian network, the probability distribution of carbon emissions under the current construction site conditions is calculated by inputting the current observation conditions. Do-calculus is used for counterfactual inference, and control interventions are implemented at each observation node. The changes in carbon emissions before and after the intervention of each factor are calculated, the marginal impact of each factor on carbon emissions is quantified, and the key factors that have a significant impact on carbon emissions are identified based on the quantification results of the marginal impact.

[0011] Furthermore, the construction of a multi-objective optimization model, combined with the key factors, to generate a low-carbon construction scheduling strategy includes: Define the state space of the multi-objective optimization model, which includes the current process queue state, equipment availability state, resource load state, predicted carbon emission factor value for future periods, and environmental condition state. Define the action space of the multi-objective optimization model, which includes adjustments to the execution sequence of work processes, selection of construction periods, adjustments to equipment configuration, and adjustments to resource allocation; A composite reward function is constructed, which integrates negative incentives from carbon emission intensity, negative incentives from construction period deviation, positive incentives from resource utilization rate, and negative incentives from conflict penalty. A multi-objective optimization model is trained based on the composite reward function, and a low-carbon construction scheduling strategy is generated by combining the quantitative analysis results of key factors affecting carbon emissions.

[0012] Secondly, embodiments of the present invention provide a low-carbon construction scheduling optimization device, comprising: The multi-source data acquisition module is used to acquire multi-source data through a multi-source heterogeneous sensor network deployed at the construction site, preprocess the multi-source data, and construct a standardized time-series dataset. The energy consumption data acquisition module is used to identify the operating status of construction equipment based on the time series dataset and generate energy consumption data for each construction equipment in each operating status. The emission power calculation module is used to obtain the corresponding dynamic carbon emission factor according to the energy type of the construction equipment, and calculate the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data. The carbon emission mapping module is used to map the equipment-level carbon emission power to the process-level and component-level based on a preset mapping relationship, and output multi-scale carbon emission data. The emission causal analysis module is used to construct a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and to use the Bayesian network to identify key factors affecting carbon emissions. The construction scheduling optimization module is used to build a multi-objective optimization model and generate a low-carbon construction scheduling strategy by combining the key factors.

[0013] Thirdly, embodiments of the present invention provide an electronic device, comprising: Memory, used to store computer programs; A processor for executing the computer program; Wherein, when the processor executes the computer program, it implements the low-carbon construction scheduling optimization method described in any of the first aspects above.

[0014] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed, implements the low-carbon construction scheduling optimization method described in any of the first aspects above.

[0015] Compared with existing technologies, the low-carbon construction scheduling optimization method provided by this invention has the following advantages: It collects multi-source data through a multi-source heterogeneous sensor network deployed at the construction site, preprocesses the multi-source data, and constructs a standardized time-series dataset; it identifies the operating status of construction equipment based on the time-series dataset and generates energy consumption data for each piece of equipment under each operating status; it obtains the corresponding dynamic carbon emission factor according to the energy type of the construction equipment, and calculates the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data; it maps the equipment-level carbon emission power to the process level and component level based on a preset mapping relationship, outputting multi-scale carbon emission data; it constructs a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and uses the Bayesian network to identify key factors affecting carbon emissions; it constructs a multi-objective optimization model with carbon emission intensity, schedule deviation, and resource utilization rate as optimization objectives, and generates a low-carbon construction scheduling strategy by combining the key factors. This invention can solve the problems of extensive carbon emission control and lack of data support for scheduling optimization in traditional construction, and realize full-link traceability and scheduling optimization of carbon emissions in construction, providing reliable technical support for the green and low-carbon development of building construction. Attached Figure Description

[0016] To more clearly illustrate the technical features of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a low-carbon construction scheduling optimization method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a low-carbon construction scheduling optimization device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0021] In a first aspect, embodiments of the present invention provide a low-carbon construction scheduling optimization method, see [link to relevant documentation]. Figure 1 This is a flowchart illustrating an embodiment of a low-carbon construction scheduling optimization method provided by the present invention.

[0022] like Figure 1 As shown, the method includes the following steps: S1: Collect multi-source data through a multi-source heterogeneous sensor network deployed at the construction site, preprocess the multi-source data, and construct a standardized time-series dataset; S2: Based on the time-series dataset, identify the operating status of the construction equipment and generate energy consumption data for each construction equipment in each operating status; S3: Obtain the corresponding dynamic carbon emission factor based on the energy type of the construction equipment, and calculate the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data; S4: Based on a preset mapping relationship, the equipment-level carbon emission power is mapped to the process level and component level, and multi-scale carbon emission data is output. S5: Construct a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and use the Bayesian network to identify the key factors affecting carbon emissions; S6: Construct a multi-objective optimization model and generate a low-carbon construction scheduling strategy by combining the key factors mentioned above.

[0023] In practice, sensors are deployed at various monitoring nodes on the construction site to collect various monitoring data related to the construction process, such as energy consumption, operating conditions, transportation, and environmental data. For example, at the power monitoring node, smart meters are installed in temporary distribution boxes or branch circuits to collect power, active / reactive power, voltage, current, and related power parameters. At the fuel / gas monitoring node, diesel flow meters, CAN bus adapters, gas flow meters, etc., are deployed to record fuel and gas consumption rates and engine operating information (speed, load percentage, etc.). At the vehicle and transportation monitoring node, vehicle GPS, fuel consumption recorders, odometers, etc., are deployed to record material transportation trajectories and energy consumption. At the environmental monitoring node, temperature, humidity, wind speed, radiation, PM2.5 sensors, etc., are deployed to collect external conditions affecting emissions and operating conditions.

[0024] The above IoT nodes transmit the monitored multi-source data to the field edge gateway via wireless transmission methods such as LoRaWAN, NB-IoT, and 4G / 5G. The edge gateway then performs preprocessing on the multi-source data, including device access and communication protocol adaptation (Modbus, CAN, MQTT, HTTP, etc.), timestamp calibration and clock synchronization, data format standardization (converting to a unified time series structure), anomaly screening, data caching and breakpoint resumption, preliminary aggregation, compression, and encryption. The edge gateway configuration ensures a stable and reliable data transmission link, effectively solving the problem of discontinuous monitoring caused by the complex network environment at the construction site.

[0025] In step S2, based on the power or flow characteristics in the time-series dataset, the operating status of the equipment is identified, and the start and end times of each operating status are marked. The transition boundary of the operating status is determined. The transition boundary refers to the time node when the construction equipment switches from one operating status to another. Using the transition boundary as the dividing point, the time-series dataset is segmented into segments to form energy consumption data segments that correspond one-to-one with a single operating status of the equipment. Time integration is performed on each energy consumption data segment to obtain the cumulative energy consumption value within each segment. Finally, structured energy consumption data containing equipment identification, operating status, time period information, and cumulative energy consumption value is generated, realizing the accurate binding of energy consumption data and equipment operating status.

[0026] In step S3, a differentiated dynamic carbon emission factor update mechanism is established. The dynamic carbon emission factor refers to the quantitative indicator of carbon emission intensity that changes dynamically with time and energy supply conditions. A time series database of carbon emission factors is pre-built. The energy type identifier, time period information and energy consumption value of the construction equipment are extracted from the energy consumption data generated in step S2. These are used as query conditions to accurately match the dynamic carbon emission factor of the corresponding energy type in the corresponding time period from the time series database of carbon emission factors. Based on the matched dynamic carbon emission factor and the cumulative energy consumption value in the energy consumption data, the equipment-level carbon emission power is calculated.

[0027] In step S4, a standardized equipment-process and process-component mapping relationship is pre-constructed to clarify the correspondence between construction equipment and construction processes, and between construction processes and building components. Based on the mapping relationship table, the equipment-level carbon emission power calculated in the above steps is transferred, aggregated and allocated level by level to finally generate multi-scale carbon emission data at the equipment level, process level and component level, so as to realize full-chain traceability of construction carbon emissions.

[0028] To accurately identify the core drivers of construction carbon emissions and quantify the marginal impact of each factor on carbon emissions, step S5 constructs a Bayesian Network (BN) causal inference model based on standardized time-series datasets and multi-scale carbon emission data. Through network inference and counterfactual analysis, the key factors affecting carbon emissions are identified.

[0029] Step S6, based on the identified key factors influencing carbon emissions, constructs a multi-objective optimization model. Through standardized state definition, action design, reward function construction, and model training, it achieves coordinated optimization of carbon emission intensity, construction period deviation, and resource utilization, generating a scientific and feasible low-carbon construction scheduling strategy, and ultimately realizing the intelligent and low-carbon upgrade of construction scheduling.

[0030] In summary, this invention, through a comprehensive technical solution encompassing data acquisition, status identification, emission calculation, multi-scale mapping, factor identification, and scheduling optimization, achieves precise accounting and traceability of construction carbon emissions across the entire chain, from equipment and process levels to component levels. It leverages Bayesian network causal inference to accurately identify key influencing factors of carbon emissions and utilizes a multi-objective optimization model to generate low-carbon construction scheduling strategies that balance reducing carbon emission intensity, controlling schedule deviations, and improving resource utilization. This significantly enhances the level of refined management of construction carbon emissions, achieves targeted emission reduction, and ensures controllable construction schedules and efficient resource utilization, providing reliable technical support for the green and low-carbon development of building construction.

[0031] In one optional implementation, the preprocessing of the multi-source data includes: The collected multi-source data is resampled to a uniform time interval; Anomaly detection is performed on resampled multi-source data to identify and mark outliers in the data; For detected abnormal and missing data, corresponding repair methods are used to repair them.

[0032] Specifically, to ensure temporal consistency among multi-source data, this embodiment resamples heterogeneous multi-source data with different acquisition frequencies and timestamps to a unified 1-minute interval, achieving strict alignment of all multi-source data in the time dimension and eliminating the time dimension heterogeneity problem caused by differences in acquisition frequencies of different sensors.

[0033] Furthermore, a hybrid detection scheme combining the sliding window ±3σ statistical threshold method and the isolated forest machine learning model is adopted for anomaly detection. First, a 5-minute sliding time window is set, and the mean and standard deviation of the data within each window are calculated. Data exceeding the mean ±3σ range are marked as suspected outliers. Then, the suspected outliers are input into the isolated forest model, which has been pre-trained using historical normal monitoring data from the construction site. The unsupervised learning capability of the model is used to further determine the abnormal attributes of the suspected outliers. Finally, the data confirmed as outliers are uniformly marked, and the collection time, data type, sensor identifier, and other related information of the outliers are recorded, so as to achieve accurate identification and traceable marking of outliers in multi-source data.

[0034] The abnormal data marked in the above steps are corrected and repaired to eliminate interference from abnormal data in subsequent analysis. For missing data generated during the transmission or acquisition of multi-source data, corresponding repair methods are adopted according to the different durations of the missing data. For short-term missing data, forward imputation combined with spline interpolation is used for completion. For long-term missing data, Kalman filtering algorithm combined with the historical operation pattern of the equipment is used for estimation and completion. After repairing long-term missing data, the quality factor corresponding to the data in that period is reduced simultaneously to reflect the confidence level of the repaired data. Through the above differentiated repair methods, the abnormal and missing data are accurately and reasonably repaired, and the true characteristics of the data are restored to the greatest extent.

[0035] Through the above-described multi-source data preprocessing process, this implementation method achieves time unification, anomaly removal, and missing data completion for multi-source heterogeneous data at the construction site. It effectively solves problems such as time dimension heterogeneity, data distortion, and data packet loss in data collected by sensor networks at the construction site. The output standardized time-series data has the technical characteristics of time consistency, data integrity, and feature authenticity, providing high-quality and reusable basic data for subsequent steps. It ensures the calculation accuracy and analysis reliability of the entire low-carbon construction scheduling optimization method from the data source.

[0036] In one optional implementation, the step of identifying the operating status of construction equipment based on the time-series dataset and generating energy consumption data for each piece of construction equipment in each operating status includes: Based on the time-series dataset, equipment operation feature parameters are extracted to identify the operating status of the construction equipment. The operating status includes at least the shutdown status, standby status, light load operating status, and heavy load operating status. The time-series dataset is segmented based on the transition boundaries of the operating state to form energy consumption data segments that correspond one-to-one with the operating state of the equipment. The cumulative energy consumption value within each energy consumption data segment is calculated to obtain the energy consumption data of each construction equipment in each operating state. The energy consumption data includes equipment identification, operating state, time period information, and cumulative energy consumption value.

[0037] Specifically, the time-series dataset is first grouped according to Equipment Identifier (EID). For different types of equipment, such as electric drive, internal combustion machinery, and transport vehicles, the corresponding equipment operation characteristic parameters are extracted from the grouped time-series dataset. Multi-dimensional feature parameters such as current waveform, vibration, and acoustics are used as inputs and fed into a CNN-LSTM hybrid model that has been trained on historical working condition data from the construction site. At the same time, the power / fuel consumption time series variation characteristics are combined for rule-based dual recognition. By setting equipment start-stop thresholds, power / fuel flow rate change rate thresholds, and minimum state duration constraints, the recognition results of the CNN-LSTM model are verified and corrected to avoid misjudgments by a single recognition method. Finally, through the combination of model recognition and rule verification, the operating status of each piece of equipment at different time periods is accurately determined. At the same time, the start time, end time, and state attributes of each operating status are marked, realizing accurate identification and traceable recording of four core states: shutdown, standby, light load operation, and heavy load operation.

[0038] For each device's time-series dataset, all operational state transition boundaries marked in the previous step are extracted. Using these boundaries as the segmentation criteria, the time-series dataset is segmented along the time axis. During the segmentation process, complete information such as timestamps, device identifiers, and feature parameters of the data in each time period is strictly preserved. After the segmentation is completed, basic attribute labels containing start time, end time, device identifier (EID), and operational state are generated for each energy consumption data segment, forming a structured energy consumption data segment. This defines a clear and independent time range and data boundary for the subsequent calculation of cumulative energy consumption values.

[0039] For each energy consumption data segment with attribute tags, the core energy consumption monitoring data within the segment is extracted according to the equipment type. For electric drive equipment, the instantaneous active power time series data is extracted; for internal combustion machinery, the fuel consumption rate time series data is extracted; and for transport vehicles, energy consumption calibration is performed by combining fuel consumption rate and mileage. Then, according to the energy consumption data type, the cumulative energy consumption value within the segment is calculated through time integration. After the calculation is completed, all energy consumption data segments of each equipment are summarized and structured, and complete core information is added to each segment. The final generated energy consumption data of construction equipment by status includes at least four core elements: equipment identifier (EID), operating status (stop / standby / light load / heavy load), time period information (start and end of the corresponding energy consumption data segment), and cumulative energy consumption value. At the same time, extended information such as energy consumption unit (unit), data quality factor (q), and energy type (type (i)) can be added according to a unified data model to form standardized energy consumption data units.

[0040] This embodiment achieves accurate and multi-dimensional identification of the operating status of construction equipment and generates energy consumption data for each piece of construction equipment in each operating state, realizing accurate binding of energy consumption data with equipment operating status, and providing refined energy consumption basic data for subsequent carbon emission power calculation.

[0041] In one optional implementation, the step of obtaining the corresponding dynamic carbon emission factor based on the energy type of the construction equipment, and calculating the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data, includes: Extract the energy type identifier, time period information, and energy consumption value of the construction equipment from the energy consumption data; Based on the energy type identifier and the time period information, the dynamic carbon emission factor of the corresponding energy type in the corresponding time period is queried from the pre-built carbon emission factor time series database; wherein, the carbon emission factor time series database is used to store the carbon emission factor values ​​of a specific energy type in a specific time period; Based on the energy consumption value and the dynamic carbon emission factor, the equipment-level carbon emission power is calculated.

[0042] It should be noted that the dynamic carbon emission factor of this invention refers to a quantitative indicator of carbon emission intensity that changes dynamically with time and energy supply conditions. Unlike the traditional fixed carbon emission factor, it can accurately reflect the actual carbon emission intensity of different time periods and different quality energy sources. The carbon emission factor time series database refers to a standardized database pre-constructed by this invention for storing the dynamic carbon emission factor values ​​of specific energy types in specific time periods. The database is classified and stored according to energy type. The core storage content includes energy type, time period range, dynamic carbon emission factor value, data source and update time. It supports quick query by energy type identifier and time period information, and also has an automatic update function to ensure the timeliness and accuracy of factor values.

[0043] To address the dynamic characteristics of different energy types at construction sites, a differentiated carbon factor update mechanism is adopted, synchronously updating the carbon emission factor time-series database to ensure the real-time accuracy of factor values ​​within the database. For electrical energy, the system automatically connects to the regional power grid dispatch platform to obtain external data such as real-time grid load curves, the proportion of clean energy (wind power, photovoltaic, hydropower, etc.), and dispatch periods. A carbon factor update model is constructed, incorporating both trend terms (long-term trend of clean energy proportion) and periodic terms (daily peak-valley differences, seasonal differences), to calculate and update the electricity carbon intensity in real time. This system enables time-series correction of the dynamic carbon emission factor for electricity, ensuring that the carbon emission factor for electricity at different times matches the actual power grid structure. For fossil fuels such as diesel, gasoline, and natural gas, the system updates the carbon emission factor according to fuel supply batches. When each batch of fuel arrives, the system records the fuel quality inspection report, supply batch number, and supply time period, determining the carbon emission factor value for that batch of fuel based on the fuel type. , Presented as a piecewise constant, the same factor value is used for the same batch of fuel during the supply period, while different batches use different factor values ​​due to quality differences, accurately reflecting the impact of fuel quality on carbon emission intensity.

[0044] Specifically, the energy type identifier, time period information, and energy consumption value of construction equipment are extracted from energy consumption data. The energy type identifier and time period information are used as query conditions and input into the carbon emission factor time series database. The system automatically matches the dynamic carbon emission factor for the corresponding energy type and time period. Based on the energy consumption value and dynamic carbon emission factor, the equipment-level carbon emission power is calculated. The calculation formula is as follows: ; in, Let be the carbon emission power of device i during time period t (i.e., the time period corresponding to the energy consumption data). Let be the cumulative energy consumption of device i during time period t. Let be the dynamic carbon emission factor of device i corresponding to the energy type during time period t.

[0045] This embodiment solves the calculation error problem caused by traditional fixed carbon emission factors. By constructing a dynamic carbon factor update mechanism and combining real-time updates and accurate queries of the carbon emission factor time series database, the carbon emission factors can accurately match the dynamic changes in energy types at the construction site. This improves the calculation accuracy of equipment-level carbon emission power at the core calculation parameter level, ensuring that emission data fits the actual construction scenario.

[0046] In one optional implementation, the step of mapping the equipment-level carbon emission power to the process-level and component-level based on a preset mapping relationship, and outputting multi-scale carbon emission data, includes: Based on the preset equipment-process mapping relationship, the carbon emission power of all related equipment in each construction process is summed by time integral to obtain process-level carbon emission data. Based on the preset process-component mapping relationship, the process-level carbon emission data is allocated to each associated component according to the weight ratio of the engineering quantity, thereby generating component-level carbon emission data; By integrating carbon emission data at the equipment, process, and component levels, multi-scale carbon emission data is formed, which includes multi-dimensional identifiers, time period information, and carbon emission values.

[0047] It should be noted that the equipment-process mapping relationship refers to the standardized mapping rules pre-constructed in this invention to clarify the correspondence between construction equipment and construction processes, i.e., the equipment-process mapping table. The core content includes equipment identifier (EID), process identifier (PID), and the time range in which the equipment participates in the corresponding process, clarifying which construction process each piece of construction equipment serves during which time period. It can be dynamically adjusted according to the construction plan to ensure that the mapping relationship is consistent with the actual construction scenario. The process-component mapping relationship refers to the standardized mapping rules used to clarify the correspondence between construction processes and building components, i.e., the process-component mapping table. The core content includes process identifier (PID), component identifier (CID), and the engineering quantity or operation weight corresponding to the component, clarifying which building components each construction process serves. It is the core basis for the transfer of process-level carbon emissions to the component level.

[0048] Specifically, the calculation formula for process-level carbon emission data is as follows: ; First, the preset equipment-process mapping table is invoked, and all construction processes are grouped according to the process identifier. For each process j, its associated equipment set is filtered out through the mapping table. Simultaneously extract the construction time window for this process. Then call all the previously calculated associated devices. carbon emission power The carbon emission power time-series data of these devices within the construction time window of process j are selected. Then, the calculation is performed according to the formula. First, the carbon emission power of all related devices within the same time period is summed to obtain the real-time total emission power of process j in time period t. Then, the real-time total emission power is integrated over time within the construction time window of process j to obtain the cumulative carbon emission of process j. After the calculation is completed, corresponding process-level carbon emission data is generated for each process, and process identifier, construction period, related equipment list and cumulative carbon emission amount are associated to ensure that process-level emission data can be traced back to specific equipment.

[0049] Furthermore, the calculation formula for component-level carbon emission data is as follows: ; First, the preset process-component mapping table is invoked, and the corresponding component set is associated with the process identifier. For each process j and its cumulative carbon emissions... Filter out all building components served by this process, and extract the quantity or work weight of each associated component k. Calculate the total workload or work weight of all components associated with the process, then calculate the weight ratio of each component k. Multiply the cumulative carbon emissions of process j by the weight ratio of each component k to obtain the cumulative carbon emissions of each component k. After the calculation is completed, corresponding component-level carbon emission data is generated for each component, and the component identifier, associated process identifier, weight ratio and cumulative carbon emission amount are associated to ensure that the component-level emission data can be traced back to the specific process and corresponding equipment.

[0050] By integrating carbon emission data at the equipment, process, and component levels, standardized multi-scale carbon emission data is formed. Equipment-level data reflects the emission details of a single piece of equipment under different conditions, process-level data reflects the overall emission level of a single process, and component-level data reflects the carbon footprint of a single building component. The three types of data are interconnected and can be queried through identification association (such as querying the emission data of its associated processes and corresponding equipment by component identification), realizing multi-scale, full-chain traceability of construction carbon emissions.

[0051] This embodiment realizes the multi-scale transmission and precise correlation of construction carbon emissions. Through the preset equipment-process and process-component mapping tables, the originally isolated equipment-level carbon emission power is aggregated into process-level emissions and distributed into component-level emissions, realizing the full-link connection of carbon emissions at the three levels of "equipment-process-component" and solving the problem that traditional emission monitoring cannot trace the emission source.

[0052] In one optional implementation, the step of constructing a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and using the Bayesian network to identify key factors affecting carbon emissions, includes: Factors related to carbon emissions are selected from the time-series dataset as observation nodes of the Bayesian network, and process-level carbon emissions are used as target nodes of the Bayesian network. The Bayesian network is constructed based on the causal relationships between the nodes. Based on the constructed Bayesian network, the probability distribution of carbon emissions under the current construction site conditions is calculated by inputting the current observation conditions. Do-calculus is used for counterfactual inference, and control interventions are implemented at each observation node. The changes in carbon emissions before and after the intervention of each factor are calculated, the marginal impact of each factor on carbon emissions is quantified, and the key factors that have a significant impact on carbon emissions are identified based on the quantification results of the marginal impact.

[0053] Specifically, potential drivers related to carbon emissions are batch-screened from time-series datasets, redundant variables unrelated to emissions are removed, the specific types and parameter ranges of observation nodes are determined, and process-level carbon emission data are extracted from multi-scale carbon emission data as target nodes of the Bayesian network. Based on the actual construction logic and engineering experience at the construction site, the causal relationships between nodes are sorted out, and the topology of the Bayesian network is constructed.

[0054] It should be noted that observation nodes refer to construction site variables that can be directly observed and measured, which are potential drivers of carbon emissions. This invention explicitly selects environmental variables (temperature, wind speed, humidity, PM2.5), personnel input, material arrival records, equipment load, and operating condition categories (shutdown, standby, light load, heavy load) at the construction site as observation nodes, covering three core dimensions: environment, construction organization, and equipment operation.

[0055] The target node refers to the core outcome variable that needs to be analyzed and predicted. This invention takes the process-level carbon emissions (taken from the process-level cumulative carbon emissions in multi-scale carbon emission data) as the core target node, because process-level emissions are the key link connecting equipment operation and component construction, and can accurately reflect the carbon emission level of the construction process.

[0056] The causal relationship between nodes refers to clarifying the "cause-effect" relationship between each observation node and the target node, and between observation nodes, based on the actual construction logic at the construction site (such as increased equipment load → increased carbon emissions in the process, excessively high temperature → decreased equipment efficiency → increased carbon emissions in the process).

[0057] Subsequently, historical monitoring data from the construction site (including historical time-series datasets and historical multi-scale carbon emission data) are used to learn the rationality of the network structure through machine learning algorithms (such as Bayesian estimation). At the same time, conditional probability tables for each node are generated to clarify the probability distribution of "parent node combination → child node". This completes the construction and calibration of the Bayesian network, ensuring that the network structure fits the actual construction and that the probability parameters are accurate and reliable.

[0058] After the Bayesian network is trained, the observation conditions of the current construction site are acquired in real time. The real-time values ​​of each observation node are input into the constructed Bayesian network. Based on the network's topology and conditional probability table, the posterior probability of carbon emissions from the process is calculated. The formula is as follows: ; in, Let Z represent the joint probability of process carbon emissions occurring simultaneously with the current observation condition Z, and let P(Z) represent the prior probability of the current observation condition Z occurring. The formula for calculating the joint probability is: ; in, It represents any node variable (including observation node and target node) at the construction site. Representing variables The causal parent node (i.e., the one that directly influences) (nodes).

[0059] Then, using the conditional expectation formula, the conditional expectation of the carbon emissions of the process under the current observation condition Z is calculated to obtain the baseline predicted value of the carbon emissions of the process under the current operating conditions. The formula is as follows: ; Where c represents a possible value for the carbon emissions of the process. This represents the posterior probability when the carbon emission value is c, and this expected value serves as the benchmark for subsequent counterfactual inferences.

[0060] The final output is the probability distribution (posterior probability distribution) and conditional expectation of carbon emissions under the current working conditions, which clarifies the approximate range and most likely value of carbon emissions under the current construction scenario, laying the foundation for subsequent counterfactual inference and marginal impact quantification.

[0061] Furthermore, do-calculus is used for counterfactual inference, selecting each observation node in the Bayesian network as the variable to be intervened. Set reasonable control intervention procedures one by one. Simulate feasible adjustment scenarios in actual construction (such as shifting from centralized material delivery to staggered delivery, reducing personnel input by 10%, adjusting heavy equipment load to light load, and controlling temperature within a suitable range). For each intervention, calculate the posterior probability distribution of carbon emissions from the process after the intervention using the posterior distribution formula. The formula is as follows: ; Then, the expected condition after intervention is calculated using the expected condition formula. Based on the expected condition after intervention and the baseline expected condition before intervention, the change in emissions at each observation node after intervention is calculated. The formula is as follows: ; Finally, the emission changes of all observation nodes were sorted, and nodes with large changes and significant impact on carbon emissions were selected as key influencing factors of construction carbon emissions. At the same time, the emission sensitivity of each key factor was quantified, the emission reduction potential of different factors was identified, and a list of key influencing factors and a marginal impact quantification report were generated to ensure that the identification results can be directly used for subsequent construction scheduling optimization.

[0062] This embodiment addresses the core pain point of traditional influencing factor analysis, which emphasizes correlation over causation. By constructing a Bayesian network causal inference model, it clarifies the causal relationship between the observed nodes (environment, personnel, materials, equipment) and the target node (process carbon emissions), making the identification results of key influencing factors more scientific and interpretable, and avoiding the misjudgment problem of traditional methods.

[0063] In one optional implementation, the step of constructing a multi-objective optimization model and generating a low-carbon construction scheduling strategy by combining the key factors includes: Define the state space of the multi-objective optimization model, which includes the current process queue state, equipment availability state, resource load state, predicted carbon emission factor value for future periods, and environmental condition state. Define the action space of the multi-objective optimization model, which includes adjustments to the execution sequence of work processes, selection of construction periods, adjustments to equipment configuration, and adjustments to resource allocation; A composite reward function is constructed, which integrates negative incentives from carbon emission intensity, negative incentives from construction period deviation, positive incentives from resource utilization rate, and negative incentives from conflict penalty. A multi-objective optimization model is trained based on the composite reward function, and a low-carbon construction scheduling strategy is generated by combining the quantitative analysis results of key factors affecting carbon emissions.

[0064] Specifically, the multi-objective optimization model takes carbon emission intensity, schedule deviation, and resource utilization rate as the core optimization objectives. First, the comprehensive state of the construction site is represented as a state vector St, which includes the current process queue state, equipment availability state, resource load state, predicted carbon emission factor value for future periods, and environmental condition state. For a given state, an action set At is defined, including scheduling strategies such as adjusting the process execution order, selecting construction time periods, adjusting equipment configuration, and adjusting resource allocation.

[0065] Combining the three core optimization objectives, a composite reward function is constructed: ; in, Let be the carbon emission intensity at time t. The schedule deviation at time t represents the difference between the actual construction progress and the planned construction progress at time t. Contribution to equipment utilization at time t Let be the conflict penalty term at time t. Let be the weight coefficients of each evaluation indicator, and satisfy . It can be dynamically adjusted according to the actual needs of the construction site.

[0066] Finally, the DQN algorithm is used to train a multi-objective optimization model. The model is trained offline in a construction simulation environment to obtain a converged construction scheduling strategy. The trained model is then deployed to the actual construction site and run online using a rolling time-domain approach to output a specific low-carbon construction scheduling strategy.

[0067] This embodiment can significantly reduce carbon emissions while ensuring construction period and equipment efficiency, and achieve multi-objective synergistic optimization of emission intensity, construction progress and equipment utilization.

[0068] Secondly, embodiments of the present invention provide a low-carbon construction scheduling optimization device, see [link to relevant documentation]. Figure 2This is a schematic diagram of an embodiment of a low-carbon construction scheduling optimization device provided by the present invention.

[0069] like Figure 2 As shown, the device includes: The multi-source data acquisition module 21 is used to acquire multi-source data through a multi-source heterogeneous sensor network deployed at the construction site, preprocess the multi-source data, and construct a standardized time-series dataset. The energy consumption data acquisition module 22 is used to identify the operating status of construction equipment based on the time series dataset and generate energy consumption data for each construction equipment in each operating status. The emission power calculation module 23 is used to obtain the corresponding dynamic carbon emission factor according to the energy type of the construction equipment, and calculate the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data. The carbon emission mapping module 24 is used to map the equipment-level carbon emission power to the process-level and component-level based on a preset mapping relationship, and output multi-scale carbon emission data. The emission causality analysis module 25 is used to construct a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and to use the Bayesian network to identify key factors affecting carbon emissions. The construction scheduling optimization module 26 is used to construct a multi-objective optimization model and generate a low-carbon construction scheduling strategy by combining the key factors.

[0070] In one optional implementation, the preprocessing of the multi-source data includes: The collected multi-source data is resampled to a uniform time interval; Anomaly detection is performed on resampled multi-source data to identify and mark outliers in the data; For detected abnormal and missing data, corresponding repair methods are used to repair them.

[0071] In one optional implementation, the step of identifying the operating status of construction equipment based on the time-series dataset and generating energy consumption data for each piece of construction equipment in each operating status includes: Based on the time-series dataset, equipment operation feature parameters are extracted to identify the operating status of the construction equipment. The operating status includes at least the shutdown status, standby status, light load operating status, and heavy load operating status. The time-series dataset is segmented based on the transition boundaries of the operating state to form energy consumption data segments that correspond one-to-one with the operating state of the equipment. The cumulative energy consumption value within each energy consumption data segment is calculated to obtain the energy consumption data of each construction equipment in each operating state. The energy consumption data includes equipment identification, operating state, time period information, and cumulative energy consumption value.

[0072] In one optional implementation, the step of obtaining the corresponding dynamic carbon emission factor based on the energy type of the construction equipment, and calculating the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data, includes: Extract the energy type identifier, time period information, and energy consumption value of the construction equipment from the energy consumption data; Based on the energy type identifier and the time period information, the dynamic carbon emission factor of the corresponding energy type in the corresponding time period is queried from the pre-built carbon emission factor time series database; wherein, the carbon emission factor time series database is used to store the carbon emission factor values ​​of a specific energy type in a specific time period; Based on the energy consumption value and the dynamic carbon emission factor, the equipment-level carbon emission power is calculated.

[0073] In one optional implementation, the step of mapping the equipment-level carbon emission power to the process-level and component-level based on a preset mapping relationship, and outputting multi-scale carbon emission data, includes: Based on the preset equipment-process mapping relationship, the carbon emission power of all related equipment in each construction process is summed by time integral to obtain process-level carbon emission data. Based on the preset process-component mapping relationship, the process-level carbon emission data is allocated to each associated component according to the weight ratio of the engineering quantity, thereby generating component-level carbon emission data; By integrating carbon emission data at the equipment, process, and component levels, multi-scale carbon emission data is formed, which includes multi-dimensional identifiers, time period information, and carbon emission values.

[0074] In one optional implementation, the step of constructing a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and using the Bayesian network to identify key factors affecting carbon emissions, includes: Factors related to carbon emissions are selected from the time-series dataset as observation nodes of the Bayesian network, and process-level carbon emissions are used as target nodes of the Bayesian network. The Bayesian network is constructed based on the causal relationships between the nodes. Based on the constructed Bayesian network, the probability distribution of carbon emissions under the current construction site conditions is calculated by inputting the current observation conditions. Do-calculus is used for counterfactual inference, and control interventions are implemented at each observation node. The changes in carbon emissions before and after the intervention of each factor are calculated, the marginal impact of each factor on carbon emissions is quantified, and the key factors that have a significant impact on carbon emissions are identified based on the quantification results of the marginal impact.

[0075] In one optional implementation, the step of constructing a multi-objective optimization model and generating a low-carbon construction scheduling strategy by combining the key factors includes: Define the state space of the multi-objective optimization model, which includes the current process queue state, equipment availability state, resource load state, predicted carbon emission factor value for future periods, and environmental condition state. Define the action space of the multi-objective optimization model, which includes adjustments to the execution sequence of work processes, selection of construction periods, adjustments to equipment configuration, and adjustments to resource allocation; A composite reward function is constructed, which integrates negative incentives from carbon emission intensity, negative incentives from construction period deviation, positive incentives from resource utilization rate, and negative incentives from conflict penalty. A multi-objective optimization model is trained based on the composite reward function, and a low-carbon construction scheduling strategy is generated by combining the quantitative analysis results of key factors affecting carbon emissions.

[0076] It should be noted that the low-carbon construction scheduling optimization device provided in this embodiment of the invention is used to execute all the process steps of the low-carbon construction scheduling optimization method in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0077] Thirdly, embodiments of the present invention provide an electronic device, see [link to previous document]. Figure 3 The diagram shown is a structural schematic of an electronic device provided in an embodiment of the present invention.

[0078] like Figure 3 As shown, the device includes: Memory 31 is used to store computer programs; Processor 32 is used to execute the computer program; When the processor 32 executes the computer program, it implements the low-carbon construction scheduling optimization method as described in any of the above embodiments.

[0079] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 31 and executed by the processor 32 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0080] The processor 32 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0081] The memory 31 can be used to store the computer programs and / or modules. The processor 32 implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory 31 and calling the data stored in the memory 31. The memory 31 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 31 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0082] It should be noted that the aforementioned electronic devices include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 3 The structural diagram is merely an example of the electronic device described above and does not constitute a limitation on the electronic device. It may include more components than shown in the diagram, or combine certain components, or use different components.

[0083] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed, implements the low-carbon construction scheduling optimization method described in any of the above embodiments.

[0084] It should be understood that the implementation of all or part of the processes in the above-described low-carbon construction scheduling optimization method can also be accomplished by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above-described low-carbon construction scheduling optimization method. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0085] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. It should be noted that, for those skilled in the art, several equivalent obvious modifications and / or equivalent substitutions can be made without departing from the technical principles of the present invention, and these obvious modifications and / or equivalent substitutions should also be considered within the scope of protection of the present invention.

Claims

1. A low-carbon construction scheduling optimization method, characterized in that, include: Multi-source data is collected by a multi-source heterogeneous sensor network deployed at the construction site, the multi-source data is preprocessed, and a standardized time-series dataset is constructed. Based on the time-series dataset, the operating status of construction equipment is identified, and energy consumption data of each piece of construction equipment in each operating status is generated. The dynamic carbon emission factor is obtained according to the energy type of the construction equipment, and the equipment-level carbon emission power is calculated by combining the dynamic carbon emission factor and the energy consumption data. Based on a preset mapping relationship, the equipment-level carbon emission power is mapped to the process level and component level, and multi-scale carbon emission data is output. A Bayesian network is constructed based on the time-series dataset and the multi-scale carbon emission data, and the Bayesian network is used to identify key factors affecting carbon emissions. A multi-objective optimization model is constructed, and a low-carbon construction scheduling strategy is generated by combining the aforementioned key factors.

2. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The preprocessing of the multi-source data includes: The collected multi-source data is resampled to a uniform time interval; Anomaly detection is performed on resampled multi-source data to identify and mark outliers in the data; For detected abnormal and missing data, corresponding repair methods are used to repair them.

3. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The step of identifying the operating status of construction equipment based on the time-series dataset and generating energy consumption data for each piece of construction equipment in each operating status includes: Based on the time-series dataset, equipment operation feature parameters are extracted to identify the operating status of the construction equipment. The operating status includes at least the shutdown status, standby status, light load operating status, and heavy load operating status. The time-series dataset is segmented based on the transition boundaries of the operating state to form energy consumption data segments that correspond one-to-one with the operating state of the equipment. The cumulative energy consumption value within each energy consumption data segment is calculated to obtain the energy consumption data of each construction equipment in each operating state. The energy consumption data includes equipment identification, operating state, time period information, and cumulative energy consumption value.

4. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The step of obtaining the corresponding dynamic carbon emission factor based on the energy type of the construction equipment, and calculating the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data, includes: Extract the energy type identifier, time period information, and energy consumption value of the construction equipment from the energy consumption data; Based on the energy type identifier and the time period information, the dynamic carbon emission factor of the corresponding energy type in the corresponding time period is queried from the pre-built carbon emission factor time series database; wherein, the carbon emission factor time series database is used to store the carbon emission factor values ​​of a specific energy type in a specific time period; Based on the energy consumption value and the dynamic carbon emission factor, the equipment-level carbon emission power is calculated.

5. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The preset mapping relationship maps the equipment-level carbon emission power to the process-level and component-level, outputting multi-scale carbon emission data, including: Based on the preset equipment-process mapping relationship, the carbon emission power of all related equipment in each construction process is summed by time integral to obtain process-level carbon emission data. Based on the preset process-component mapping relationship, the process-level carbon emission data is allocated to each associated component according to the weight ratio of the engineering quantity, thereby generating component-level carbon emission data; By integrating carbon emission data at the equipment, process, and component levels, multi-scale carbon emission data is formed, which includes multi-dimensional identifiers, time period information, and carbon emission values.

6. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The process involves constructing a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and using the Bayesian network to identify key factors influencing carbon emissions, including: Factors related to carbon emissions are selected from the time-series dataset as observation nodes of the Bayesian network, and process-level carbon emissions are used as target nodes of the Bayesian network. The Bayesian network is constructed based on the causal relationships between the nodes. Based on the constructed Bayesian network, the probability distribution of carbon emissions under the current construction site conditions is calculated by inputting the current observation conditions. Do-calculus is used for counterfactual inference, and control interventions are implemented at each observation node. The changes in carbon emissions before and after the intervention of each factor are calculated, the marginal impact of each factor on carbon emissions is quantified, and the key factors that have a significant impact on carbon emissions are identified based on the quantification results of the marginal impact.

7. The low-carbon construction scheduling optimization method as described in claim 1, characterized in that, The construction of a multi-objective optimization model, combined with the key factors, to generate a low-carbon construction scheduling strategy includes: Define the state space of the multi-objective optimization model, which includes the current process queue state, equipment availability state, resource load state, predicted carbon emission factor value for future periods, and environmental condition state. Define the action space of the multi-objective optimization model, which includes adjustments to the execution sequence of work processes, selection of construction periods, adjustments to equipment configuration, and adjustments to resource allocation; A composite reward function is constructed, which integrates negative incentives from carbon emission intensity, negative incentives from construction period deviation, positive incentives from resource utilization rate, and negative incentives from conflict penalty. A multi-objective optimization model is trained based on the composite reward function, and a low-carbon construction scheduling strategy is generated by combining the quantitative analysis results of key factors affecting carbon emissions.

8. A low-carbon construction scheduling optimization device, characterized in that, include: The multi-source data acquisition module is used to acquire multi-source data through a multi-source heterogeneous sensor network deployed at the construction site, preprocess the multi-source data, and construct a standardized time-series dataset. The energy consumption data acquisition module is used to identify the operating status of construction equipment based on the time series dataset and generate energy consumption data for each construction equipment in each operating status. The emission power calculation module is used to obtain the corresponding dynamic carbon emission factor according to the energy type of the construction equipment, and calculate the equipment-level carbon emission power by combining the dynamic carbon emission factor and the energy consumption data. The carbon emission mapping module is used to map the equipment-level carbon emission power to the process-level and component-level based on a preset mapping relationship, and output multi-scale carbon emission data. The emission causal analysis module is used to construct a Bayesian network based on the time-series dataset and the multi-scale carbon emission data, and to use the Bayesian network to identify key factors affecting carbon emissions. The construction scheduling optimization module is used to build a multi-objective optimization model and generate a low-carbon construction scheduling strategy by combining the key factors.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program; Wherein, when the processor executes the computer program, it implements the low-carbon construction scheduling optimization method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed, implements the low-carbon construction scheduling optimization method as described in any one of claims 1 to 7.

Images