BIM-based carbon emission optimization method for green construction process of substation

Abstract

The application relates to the technical field of data processing, and discloses a substation green construction process carbon emission optimization method based on BIM. The method comprises the following steps: decomposing a construction process into three-level process units based on a BIM process correlation model to obtain a process node list; calculating the dynamic carbon emission intensity factor of each three-level process unit, and marking a process that continuously exceeds the upper limit of a carbon emission benchmark fingerprint bandwidth as a problem process; injecting a candidate alternative process into a real-time mirror of a concurrent state, calculating a concurrent power interference coefficient to predict carbon emission; and calculating net carbon benefits, and executing process node replacement or updating a concurrent applicability label and the upper limit of the carbon emission benchmark fingerprint bandwidth based on whether the net carbon benefits are greater than zero. The application realizes dynamic identification and minimum disturbance autonomous correction of construction process carbon emission.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method for optimizing carbon emissions during the green construction process of substations based on BIM. Background Technology

[0002] Building Information Modeling (BIM) technology, by digitizing all elements of a building project, enables visualized management of the construction process and refined extraction of quantities, and has been widely applied in the construction engineering field. In substation engineering, existing technologies are typically based on 4D-BIM, incorporating the construction schedule time dimension into the 3D model. After extracting the bill of quantities for each construction stage, multiplying it by the rated power of the corresponding machinery and the carbon emission factor, the carbon emissions for each stage are calculated. Before construction, a static planning process is performed based on a multi-objective optimization framework of time, cost, and carbon emissions, minimizing the total carbon emissions by pre-selecting the execution mode for each process. Some studies further integrate IoT sensors with the BIM platform to visualize the deviation between actual and estimated carbon emissions at the construction site, allowing project managers to manually decide whether to adjust the construction plan based on this deviation information.

[0003] However, the aforementioned existing technologies have the following shortcomings: First, carbon emission calculations always rely on the rated power of the machinery nameplate as the calculation input, which cannot reflect the actual energy consumption of construction machinery under intermittent operation and load rate variation, resulting in systematic bias in the carbon emission calculation results; Second, the existing multi-objective optimization framework uses static preset values ​​for the carbon emission parameters of each process execution mode, without considering the unique engineering scenario of highly concurrent multi-disciplinary construction of substations, primary equipment, secondary equipment, and cable construction, where the competitive occupation of temporary distribution branch capacity by concurrent processes leads to limited actual available power of candidate alternative processes, thus causing their true carbon emissions to deviate significantly from the preset value; Third, bias perception relies on human intervention, and when the concurrent state of multiple disciplines continues to change dynamically, the response speed and judgment accuracy of human decision-making are difficult to meet the needs of real-time optimization on the construction site.

[0004] If measured power is introduced to replace rated power to correct carbon emission calculation deviations, a dynamic judgment benchmark that can distinguish between normal operating condition fluctuations and structural deviations must be established. Single-point thresholds alone cannot reliably identify continuous carbon emission anomalies caused by concurrent interference. Furthermore, if the execution decision of candidate replacement processes is shifted from pre-static planning to dynamic triggering during the construction execution phase, a new problem arises: the actual carbon emissions of candidate replacement processes under the current multi-disciplinary concurrent state are unknown. The limitations of static preset factors are further amplified at this stage. When a virtual pre-execution mechanism is introduced to predict the dynamic carbon emissions of candidate replacement processes in real time, whether the carbon emission costs of mechanical movement, shutdown and startup, and preparation work generated by the replacement behavior itself exceed the carbon benefits that the replacement can bring constitute a new problem to be solved. If the replacement cost is not included in the decision boundary, blindly executing the replacement operation will introduce new net carbon losses, making the optimization effect counterproductive. Summary of the Invention

[0005] This application provides a BIM-based method for optimizing carbon emissions during the green construction process of substations. It solves the problems that the actual carbon emissions of candidate alternative procedures are unknown due to competition for power distribution capacity in multi-disciplinary concurrent construction scenarios of substations, and that the carbon emission cost of the substitution behavior itself is not included in the decision boundary, leading to optimization failure. It realizes dynamic identification and minimum disturbance autonomous correction of carbon emissions during the construction process.

[0006] This application provides a BIM-based method for optimizing carbon emissions during the green construction process of substations. The BIM-based method for optimizing carbon emissions during the green construction process of substations includes: Step S1: Based on the BIM process association model of the substation project, the substation construction process is decomposed into several third-level process units to obtain a list of process nodes. Step S2: Based on the measured power and carbon emission factor of the machinery corresponding to each level 3 process unit, calculate the dynamic carbon emission intensity factor of each level 3 process unit, and mark the level 3 process units whose dynamic carbon emission intensity factor continuously exceeds the upper limit of the carbon emission benchmark fingerprint bandwidth as problem processes. Step S3: Inject the candidate replacement process of the problem process into the real-time concurrent state mirror carrying the current measured power, spatial coordinates and remaining available capacity of the distribution branch of each third-level process unit. Calculate the concurrent power interference coefficient based on the ratio of the rated power of the candidate replacement process to the remaining available capacity of the distribution branch. Substitute the concurrent power interference coefficient into the carbon emission prediction formula of the candidate replacement process to obtain the predicted carbon emission of the candidate replacement process under the current concurrent state. Step S4: Calculate the net carbon gain based on the remaining workload of the problematic process, the dynamic carbon emission intensity factor, the predicted carbon emissions, and the carbon emissions from process replacement. Using the net carbon gain being greater than zero as a criterion, replace the node of the problematic process in the BIM process association model with the candidate alternative process, or write the current concurrent state feature vector into the concurrent applicability label of the candidate alternative process and update the upper bound of the carbon emission benchmark fingerprint bandwidth.

[0007] The technical solution provided in this application introduces a dynamic carbon emission intensity factor to replace the static rated carbon emission factor in the existing technology during the substation construction execution phase. This transforms the input for carbon emission calculation from the rated power on the mechanical nameplate to the periodically averaged measured power based on measured current, line voltage, and fuel flow. This fundamentally eliminates the systematic calculation bias caused by the neglect of intermittent operations and load rate changes in the existing technology. The carbon emission benchmark fingerprint bandwidth is used to construct an interval judgment model with the mean plus or minus two standard deviations as the boundary. This distinguishes between carbon emission fluctuations under normal operating conditions and structural offsets caused by concurrent interference. The continuous judgment mechanism for ten consecutive collection cycles further eliminates false triggers caused by single load peaks, making the identification results of problematic procedures physically interpretable, rather than simply threshold exceeding alarms. After injecting the candidate alternative process into the real-time concurrent state mirror carrying real-time concurrent state data, the actual power and duration of the correction of the candidate alternative process are adjusted in conjunction with the concurrent power interference coefficient. This enables dynamic prediction of the actual carbon emissions of the same alternative process under different multi-disciplinary concurrent states, overcoming the fundamental defect of the existing time-cost-carbon emission multi-objective optimization framework that uses static preset factors to replace the measured state and cannot capture the concurrent disturbance effect.

[0008] The calculation of net carbon benefit incorporates the carbon emissions from process replacement as an independent variable into the decision boundary, including the sum of carbon emissions from mechanical movement, shutdown and startup, and preparation work. This upgrades the basis for replacement decisions from a simple comparison of process carbon emissions to a complete benefit calculation that includes the cost of the replacement itself, avoiding the introduction of new net carbon losses due to performing replacement operations with negative carbon benefits. When the net carbon benefit is not greater than zero, the current concurrent state feature vector is written into the concurrent applicability label of the candidate replacement process. The label data that accumulates continuously as construction progresses allows the process library to learn the actual replacement effectiveness of each candidate replacement process under different concurrent states. Candidate processes with a cumulative successful replacement rate of less than one-fifth are removed, reducing the number of subsequent invalid virtual verification calculations. The temporary adaptive correction of the upper bound of the carbon emission benchmark fingerprint bandwidth avoids repeatedly triggering replacement processes destined to fail during the continuous existence of the same concurrent interference. After the five-minute effective window ends, the original bandwidth is automatically restored, ensuring that the judgment mechanism returns to normal sensitivity after the concurrent state changes. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of an embodiment of the carbon emission optimization method for green construction process of substation based on BIM in this application. Figure 2 This is a schematic diagram of the real-time monitoring of dynamic carbon emission intensity factor and the determination of problematic processes in an embodiment of this application. Detailed Implementation

[0011] This application provides a BIM-based method for optimizing carbon emissions during the green construction process of substations. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0012] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the carbon emission optimization method for green construction process of substations based on BIM in this application includes: Step S1: Based on the BIM process association model of the substation project, the substation construction process is decomposed into several third-level process units to obtain a list of process nodes. Specifically, the BIM process association model is a 3D information model formed by inputting five categories of professional components—civil structure, primary power equipment, secondary power equipment, cable lines, and temporary facilities—into BIM modeling software based on the substation engineering design drawings. Each component object carries five attributes: component code, material type, design quantity, process number, and planned start and end time nodes. A third-level process unit is the smallest executable construction unit defined below the first and second-level processes, with the completion of a single construction result by a single piece or group of construction machinery as the boundary. For example, the concrete pouring of a pile cap under a pile foundation project constitutes a third-level process unit. The process node list is a structured record set formed by extracting each of the above three-level process units. Each record carries three initial parameters: the type of machinery called, the rated power of the machinery, and the planned quantity, and maintains a bidirectional index relationship with the BIM process association model through the component code. The carbon emission benchmark fingerprint bandwidth is a normal operating condition judgment interval determined by calculating the mean and standard deviation of the historical observation values ​​of carbon emission intensity factors of the same three-level process units based on historical construction data of no less than three similar substations. The normal operating condition judgment interval is determined by using the mean plus or minus two standard deviations as the boundary. Two standard deviations correspond to a historical observation coverage of 95.45%. The basis for selecting this confidence range is that the load rate of substation construction machinery approximately follows a normal distribution under normal operating conditions. Observations that deviate from this interval are considered to be affected by abnormal factors.

[0013] Step S2: Based on the measured power and carbon emission factor of the machinery corresponding to each level 3 process unit, calculate the dynamic carbon emission intensity factor of each level 3 process unit, and mark the level 3 process units whose dynamic carbon emission intensity factor continuously exceeds the upper limit of the carbon emission benchmark fingerprint bandwidth as problem processes. Specifically, the dynamic carbon emission intensity factor is the carbon emission intensity value per unit of work volume obtained by converting the actual energy consumption of the machinery corresponding to the current three-level process unit into carbon emissions within a collection cycle, and then dividing it by the amount of work completed within the same cycle. Unlike the existing technology that uses the rated power of machinery instead of the measured power to calculate the carbon emission factor, this scheme calculates the measured power of electric machinery by multiplying the actual power supply current, line voltage, and power factor, and calculates the measured power of fuel machinery by multiplying the measured fuel flow rate, diesel density, and lower heating value. The collection cycle is 30 seconds, and multiple samples are taken within the cycle to calculate the average value, ensuring that the actual impact of intermittent operation and load rate changes on carbon emissions is captured. The out-of-range judgment for ten consecutive collection cycles, i.e., five minutes, is set to exclude single occasional peak interference and confirm that the machinery has a continuous carbon emission deviation under the influence of concurrent load before triggering the subsequent process.

[0014] Step S3: Inject the candidate replacement process of the problem process into the real-time concurrent state mirror carrying the current measured power, spatial coordinates and remaining available capacity of the distribution branch of each third-level process unit. Calculate the concurrent power interference coefficient based on the ratio of the rated power of the candidate replacement process to the remaining available capacity of the distribution branch. Substitute the concurrent power interference coefficient into the carbon emission prediction formula of the candidate replacement process to obtain the predicted carbon emission of the candidate replacement process under the current concurrent state. Specifically, the concurrent real-time mirror is a dynamic data carrier formed by continuously writing the cycle-average measured power of the machinery corresponding to each of the three-level process units at the construction site, the three-dimensional spatial coordinates of the machinery output by the ultra-wideband positioning tag, and the remaining available power capacity collected by the smart meters of each power distribution branch into the corresponding process node of the BIM process association model with an update cycle of 30 seconds and the process node number as the index key. Unlike the existing 4D-BIM technology, which only records the construction progress results, the concurrent real-time mirror completely records the process state quantities that determine the carbon emission results. The concurrent power interference coefficient is a quantitative characterization of the ratio of the actual available power of a candidate replacement process to its rated power under the current load condition of the distribution branch. When the remaining available power capacity of the distribution branch is not lower than the rated power of the candidate replacement process, the coefficient is taken as one, reflecting that the candidate replacement process can operate under rated conditions. When the remaining available power capacity is lower than the rated power, the coefficient is taken as the ratio of the two, reflecting the degree of compression of the actual available power of the candidate replacement process caused by the competition of temporary power capacity due to the concurrent construction of multiple disciplines, which in turn affects the calculation results of its correction duration and predicted carbon emissions. This is a concurrent disturbance effect that cannot be captured by the existing time-cost-carbon emission multi-objective optimization framework that uses static preset carbon emission factors to replace the measured state.

[0015] Step S4: Calculate the net carbon gain based on the remaining work volume of the problem process, the dynamic carbon emission intensity factor, the predicted carbon emission, and the carbon emission from process replacement. Using whether the net carbon gain is greater than zero as a criterion, replace the node of the problem process with a candidate alternative process in the BIM process association model, or write the current concurrent state feature vector into the concurrent applicability label of the candidate alternative process and update the upper bound of the carbon emission benchmark fingerprint bandwidth.

[0016] Specifically, the calculation of net carbon benefit uses the carbon emissions of the problem process obtained by multiplying the current dynamic carbon emission intensity factor by the remaining work volume as the subtraction benchmark, and subtracts the sum of the predicted carbon emissions of the candidate replacement process and the carbon emissions of the process replacement. The carbon emissions of the process replacement include the carbon emissions generated by moving the machinery from its current position to the work position of the problem process, the fixed carbon emissions from the machinery shutdown and start-up, and the carbon emission quota value of the preparation work standard. Incorporating the carbon emission cost of the replacement behavior itself into the decision calculation boundary is the fundamental difference from the existing technology that only compares the difference in the total carbon emissions of the process. When the net carbon gain is not greater than zero, the remaining available power capacity of the current distribution branch, the spatial coordinates of each third-level process unit, and the concurrent power interference coefficient of each candidate replacement process together constitute the current concurrent state feature vector, which is written into the concurrent applicability label of the candidate replacement process. The concurrent applicability label is a dynamic attribute field that records the history of the replacement effectiveness of the candidate replacement process under different concurrent state conditions. It is continuously accumulated as construction progresses. When the ratio of the number of successful replacements of a candidate replacement process to the cumulative number of verifications is less than one-fifth, it is marked as inapplicable and removed from the candidate set to avoid repeatedly triggering invalid virtual verification calculations.

[0017] In one specific embodiment, step S1 includes: Based on the substation engineering design drawings, a three-dimensional information model containing five categories of professional components—civil structure, primary power equipment, secondary power equipment, cable lines, and temporary facilities—is established in BIM modeling software. Each component is assigned a component code, material type, design quantity, process number, and planned start and end time nodes for construction, thus obtaining a BIM process association model. Based on the process number of each component in the BIM process association model, the substation construction process is decomposed into first-level process, second-level process and third-level process in sequence. The third-level process is the smallest process unit. For each third-level process unit, process nodes carrying the type of machinery to be called, the rated power of the machinery and the planned amount of work are created to obtain the process node list. Based on the calling machine type of each process node in the process node list, candidate alternative processes that meet the construction result equivalence constraints, schedule constraints and resource availability constraints are retrieved from the BIM process library. Candidate alternative process indexes are established for the process nodes of each third-level process unit. The candidate alternative process indexes are written into the corresponding process nodes in the process node list to obtain the updated process node list carrying the candidate alternative process indexes. Based on the historical sample set of carbon emission intensity factors for each third-level process unit in the historical construction data of similar substations, the mean and standard deviation of the historical sample set of carbon emission intensity factors for each third-level process unit are calculated. The lower bound is the mean minus twice the standard deviation, and the upper bound is the mean plus twice the standard deviation. The carbon emission benchmark fingerprint bandwidth for each third-level process unit is obtained, and the carbon emission benchmark fingerprint bandwidth is written into the corresponding process node in the updated process node list.

[0018] Specifically, among the five categories of professional components in the BIM process association model, civil engineering structures include inherent civil engineering units of substations such as pile foundations, pile caps, main transformer foundations, GIS foundations, and frame foundations; primary power equipment includes primary-side equipment with voltage level attributes such as main transformers, circuit breakers, disconnect switches, current transformers, and surge arresters; secondary power equipment includes secondary-side equipment such as protection panels, control panels, and communication equipment; cable lines include power cables, control cables, and optical cables; and temporary facilities include construction enclosures, temporary roads, and temporary power distribution facilities. The design quantity field for each type of component is filled in according to the actual unit of measurement for that component. For civil engineering components, the volume (cubic meters) or mass (tons) is used; for cable lines, the length (meters) is used; and for equipment, the number of units is used. The division of tertiary process units is based on the completion of a single component construction result by a single machine or a group of collaborative machines. Concrete pouring and rebar binding of the same pile cap belong to two independent tertiary process units. The component codes of the two are the same, but the process numbers are different. They are distinguished by the process number rather than the component code. Among the three constraints of the candidate alternative process, the construction result equivalence constraint requires that the target component code of the alternative process is completely consistent with that of the problem process, the schedule constraint requires that the planned completion time of the alternative process does not exceed 120% of the remaining planned time of the problem process, and the resource availability constraint requires that the machinery called by the alternative process is currently unoccupied in the real-time mirror of the concurrent state. All three constraints must be met simultaneously to be included in the candidate alternative process index.

[0019] The basis for using plus or minus two standard deviations of the mean in the carbon emission benchmark fingerprint bandwidth is that the actual load rate distribution of substation construction machinery under normal operating conditions approximately follows a normal distribution. Two standard deviations cover 95.45% of historical normal observation samples. Observations exceeding this range are identified as abnormal carbon emission deviations caused by abnormal concurrent interference. If the value is too narrow (e.g., one standard deviation), a large number of normal fluctuations will be misjudged as problematic processes, frequently triggering invalid virtual verifications. If the value is too wide (e.g., three standard deviations), the true carbon emission deviation will be masked, and the replacement process cannot be triggered in a timely manner. Two standard deviations strike a balance between underreporting and false reporting. The historical sample set must come from no fewer than three similar substation projects. The criteria for "similar" are the same voltage level and similar construction scale. If the number of samples is less than three, plus or minus 30% of the corresponding machinery shift carbon emission quota value in Appendix A of the "Building Carbon Emission Calculation Standard" GB / T 51366-2019 is used as the bandwidth replacement value. The 30% replacement bandwidth is derived from the statistical summarization of the systematic deviation between the quota value and the measured value. When the carbon emission benchmark fingerprint bandwidth is written into the updated process node list, the bandwidth field of the corresponding process node is updated by overwriting, and the original field value is not retained, ensuring that the benchmark bandwidth is recalculated based on the latest historical data before each construction.

[0020] In one specific embodiment, step S2 includes: With a data acquisition cycle of 30 seconds, the measured current and line voltage of the power supply lines of the electric machinery corresponding to each third-level process unit are multiplied, and the measured fuel flow rate and diesel calorific value of the fuel machinery outlet pipe are multiplied to obtain the cycle-average measured power of the machinery corresponding to each third-level process unit. Multiply the periodic average measured power by the regional power grid carbon emission factor or diesel carbon emission factor, and then divide by the amount of work completed by the corresponding third-level process unit within the collection period to obtain the dynamic carbon emission intensity factor of each third-level process unit. The dynamic carbon emission intensity factor is compared with the upper bound of the carbon emission benchmark fingerprint bandwidth carried by the corresponding process node in the process node list. The third-level process unit whose dynamic carbon emission intensity factor exceeds the upper bound of the carbon emission benchmark fingerprint bandwidth in ten consecutive collection cycles is marked as a problem process, and the problem process marking result is obtained. Based on the problem process marking results, from the candidate alternative process index of the corresponding process node in the process node list, candidate alternative processes are selected where the corresponding machine is currently unoccupied and the planned completion time does not exceed 120% of the remaining planned time of the problem process, thus obtaining a set of candidate alternative processes.

[0021] Specifically, the calculation of the periodic average measured power employs different paths for electric machinery and fuel-fired machinery. For electric machinery, the measured current of the power supply line is collected using a clamp-on current transformer, and the line voltage is read using a smart meter. The measured current, line voltage, and power factor are multiplied together and then multiplied by the square root of 3 to obtain the three-phase measured power. The power factor is taken as 0.85, which is derived from the statistical range of typical power factors of commonly used induction motors near rated load in substation construction. A value that is too low will systematically overestimate the power, while a value that is too high will underestimate the power. 0.85 is at the median level of this type of machinery under rated operating conditions. For fuel-fired machinery, the fuel flow rate in the oil pipe is collected using a Hall effect flow sensor. The measured flow rate is multiplied by the diesel density and lower heating value and then divided by 3600 to obtain the real-time power of the fuel-fired machinery. The diesel density is taken as 0.85 kg / L, and the lower heating value is taken as 42,700 kJ / kg. Both parameters are taken from the appendix data of the "Building Carbon Emission Calculation Standard" GB / T 51366-2019. The 30-second data acquisition cycle is set based on the fact that the minimum significant change cycle of the load rate of substation construction machinery is usually no less than 15 seconds. A 30-second cycle captures load fluctuations while avoiding excessive sampling density that could overload edge computing nodes. The arithmetic mean of the original sampled values ​​within each cycle is used to obtain the cycle-averaged measured power, eliminating interference from single-sample spikes. The regional power grid carbon emission factor is selected based on the 2023 edition of the "China Regional Power Grid Baseline Emission Factor," choosing the corresponding power grid zone value for the project area. The diesel carbon emission factor is set at 3.096 kg CO2 equivalent per liter, a value derived from the appendix of the same national standard.

[0022] The judgment window of ten consecutive acquisition cycles, or five minutes in total, is designed to distinguish between temporary carbon emission exceedances caused by a single load peak and structural carbon emission shifts caused by the continuous compression of available mechanical power by multiple disciplines. Judgments with fewer than ten cycles will misclassify normal load fluctuations as problematic processes, while judgments with more than ten cycles will delay the response to actual carbon emission shifts. The five-minute window matches the minimum response granularity of process scheduling instructions in substation construction. The starting point for calculating when the dynamic carbon emission intensity factor exceeds the upper limit of the carbon emission benchmark fingerprint bandwidth is the current acquisition cycle as the first counting cycle. If the dynamic carbon emission intensity factor falls back to within the upper limit of the bandwidth in any cycle during continuous counting, the count is reset to zero and restarted, ensuring that the marking results reflect continuous shifts rather than intermittent fluctuations. In the screening of the candidate alternative process set, the data source for determining the unoccupied state is the current state field of the corresponding mechanical process node in the real-time mirror of the concurrent state. This field is updated along with the mechanical three-dimensional space coordinates every 30-second update cycle. The 120% schedule constraint upper limit is based on the remaining planned duration of the problematic process at the current moment, allowing the alternative process to consume 20% more time than originally planned. This tolerance range comes from the generally accepted upper limit for the schedule deviation of sub-projects in substation construction progress management. Alternative processes exceeding this range will have a chain reaction on the start time nodes of subsequent related processes.

[0023] Figure 2 This is a schematic diagram of the real-time monitoring of dynamic carbon emission intensity factor and the determination of problematic processes in an embodiment of this application. Figure 2 In the diagram, the horizontal axis represents the data collection cycle number, with each cycle corresponding to 30 seconds. The vertical axis represents the carbon emission intensity factor, expressed in kilograms of carbon dioxide equivalent per unit of project volume. The gray-filled area represents the carbon emission benchmark fingerprint bandwidth calculated based on historical construction data of similar substations. The upper limit of the bandwidth is determined by the mean plus twice the standard deviation. The black dashed line represents the static rated carbon emission intensity factor calculated using the rated mechanical power in existing technologies. The gray solid line represents the dynamic carbon emission intensity factor calculated in this scheme based on the average measured power of the mechanical cycle corresponding to each third-level process unit. It can be seen that the dynamic factor continuously increases from the 30th to the 55th data collection cycle due to the concurrent processes of multiple disciplines occupying the distribution branch capacity. From the 38th to the 48th data collection cycle, it exceeds the upper limit of the carbon emission benchmark fingerprint bandwidth for ten consecutive data collection cycles, triggering the marking of problematic processes and entering the virtual pre-execution process of candidate replacement processes. After the 48th cycle, the process replacement is completed, and the dynamic factor falls back to the bandwidth range.

[0024] In one specific embodiment, step S3 involves injecting the candidate replacement process for the problematic process into a concurrent real-time mirror image carrying the current measured power, spatial coordinates, and remaining available capacitance of each third-level process unit, including: With an update cycle of 30 seconds, the average measured power of the machinery corresponding to each level-three process unit, the three-dimensional spatial coordinates of the machinery obtained by the ultra-wideband positioning tag, and the remaining available power capacity collected by the smart meters of each power distribution branch on the construction site are written into the corresponding process node of the BIM process association model with the process node number as the index key to obtain a real-time mirror of the concurrent state. Based on the process node number of each candidate alternative process in the candidate alternative process set, the remaining available power capacity of each power distribution branch within the corresponding work space of the candidate alternative process is read from the concurrent state real-time image. The Euclidean distance between the mechanical three-dimensional space coordinates of the candidate alternative process and the work position coordinates of the problem process is written into the corresponding process node in the concurrent state real-time image, thus obtaining the concurrent state real-time image carrying the mechanical movement distance.

[0025] Specifically, the concurrent real-time mirroring is a dynamic data structure based on the BIM process association model, indexed by process node numbers, and refreshed every 30 seconds. Each process node has three types of real-time fields: the average measured power of the corresponding machinery in the current cycle, the three-dimensional spatial coordinates of the machinery output by the ultra-wideband positioning system, and the remaining available power capacity converted from the current readings of the smart meters of each power distribution branch within the machinery's operating space. All three types of fields are refreshed by overwriting, without retaining historical cycle values. Ultra-wideband positioning tags are installed on the fixed structural components of each construction machine, with a positioning accuracy of no less than 30 centimeters. The coordinate system uses the same construction coordinate origin as the BIM process association model. Coordinate alignment is calibrated once before construction using measurement control points, and no recalibration is required for each cycle after calibration. The work space range corresponding to the candidate alternative process is determined by extending 2 meters outward from the BIM geometric boundary frame of the target component of the candidate alternative process. The minimum remaining available power capacity of all power distribution branch nodes within the extended range is taken as the effective power distribution constraint value of the candidate alternative process. The basis for selecting the minimum value is that the candidate alternative process is constrained by the most strained power distribution branch capacity in the work space when it is actually connected.

[0026] The mechanical movement distance is calculated using three-dimensional Euclidean distance. The movement distance is calculated by taking the square root of the sum of the squares of the differences in the three coordinate axes between the current three-dimensional spatial coordinates of the candidate replacement process and the work position coordinates of the problem process (using the BIM geometric center point coordinates of the target component of the problem process) as the two endpoints. Three-dimensional distance is used instead of two-dimensional planar distance because substation construction sites often involve mechanical relocation across floors or elevations; calculating only the horizontal distance would underestimate the actual movement path of crawler cranes and other machines requiring vertical displacement. After the movement distance is calculated, it is appended to the corresponding process node in the concurrent real-time mirror carrying the mechanical movement distance. This field is stored alongside existing fields such as the periodic average measured power, three-dimensional spatial coordinates, and remaining available capacity of the distribution branch. The timestamp is bound to the start time of the current 30-second update cycle to ensure time alignment of fields within the same process node. The remaining available capacity of the distribution branch within the work space corresponding to the candidate replacement process is read directly after locating it in the concurrent real-time mirror using the process node number as the search key, without triggering a new round of sensor acquisition. The read value is the latest reported value from the smart meter within the current update cycle.

[0027] In one specific embodiment, step S3, calculating the concurrent power interference coefficient based on the ratio of the rated power of the candidate replacement process to the remaining available capacitance of the distribution branch, includes: Based on the real-time mirroring of the concurrent state of the carrying machinery movement distance, the rated power of the candidate replacement process is compared with the remaining available power capacity of the power distribution branch within the corresponding work space of the candidate replacement process. When the rated power is not greater than the remaining available power capacity, the concurrent power interference coefficient is recorded as one; when the rated power is greater than the remaining available power capacity, the ratio of the remaining available power capacity to the rated power is recorded as the concurrent power interference coefficient, thus obtaining the concurrent power interference coefficient of each candidate replacement process.

[0028] Specifically, the concurrent power interference coefficient is a quantitative expression of the proportion of a candidate replacement process whose rated power can be satisfied under the constraint of the actual available power capacity of the current distribution branch. Its value ranges from greater than zero to no greater than one. The rated power originates from the mechanical rated power field carried by the corresponding candidate replacement process node in the process node list. This field was written when the process node was established in step S1, and its value is based on the nameplate rated value in the mechanical equipment technical manual. The remaining available power capacity of the distribution branch is taken as the minimum remaining available capacity of all distribution branches within the corresponding work space range of the candidate replacement process in the real-time mirror of the concurrent state carrying the mechanical movement distance. This minimum value was determined and stored in the corresponding process node field in the previous step, and is directly read here without recalculation. When the rated power of the candidate replacement process is not greater than this minimum remaining available power capacity, it indicates that the current concurrent process does not effectively compress the distribution branch of the candidate replacement process, and the candidate replacement process can operate completely at its rated power. The concurrent power interference coefficient is recorded as one. When the rated power is greater than the minimum remaining available power capacity, it indicates that other concurrent processes have occupied part of the power distribution capacity, resulting in limited actual access power for the candidate replacement process. The ratio obtained by dividing the remaining available power capacity by the rated power is the proportion of the actual available power of the candidate replacement process to the rated power under the current concurrent state. This ratio is the concurrent power interference coefficient.

[0029] The concurrent power interference coefficients of multiple candidate processes in the candidate replacement process set are calculated independently for each process. When calculating the concurrent power interference coefficient of a particular candidate process, the resource status of other candidate processes is not included in the deduction of the remaining available capacity of the current distribution branch. This is because each candidate process in the candidate replacement process set is in a waiting state rather than a simultaneous execution state. Ultimately, only the candidate process with the largest net carbon benefit is executed, and there is no situation where multiple candidate processes occupy the distribution capacity simultaneously. After the concurrent power interference coefficient is calculated, it is stored in the newly added field of the corresponding candidate replacement process process node in the real-time concurrent state mirror carrying mechanical movement distance. The field name is bound to the process node number and remains unchanged within the same 30-second update cycle. When the next update cycle is triggered, it is recalculated when the remaining available capacity field of the distribution branch is refreshed, ensuring that the concurrent power interference coefficient always reflects the latest concurrent state.

[0030] In one specific embodiment, step S3 involves substituting the concurrent power interference coefficient into the carbon emission prediction formula for the candidate replacement process to obtain the predicted carbon emissions of the candidate replacement process under the current concurrent state, including: Multiply the concurrent power interference coefficient by the rated power of the candidate replacement process to obtain the corrected actual power of the candidate replacement process under the current concurrent state. Divide the remaining workload of the problematic process by the product of the unit time productivity of the candidate alternative process under rated power and the concurrent power interference coefficient to obtain the correction duration of the candidate alternative process under the current concurrent state. Multiply the corrected actual power by the regional power grid carbon emission factor and the correction duration, and add the product of the rated fuel consumption and concurrent power interference coefficient of the fuel machinery of the candidate alternative process, the diesel carbon emission factor and the correction duration to obtain the predicted carbon emissions of the candidate alternative process under the current concurrent state.

[0031] Specifically, the corrected actual power is the value obtained by multiplying the concurrent power interference coefficient by the rated power of the candidate replacement process. It reflects the actual power level that the electric machinery of the candidate replacement process can access and operate stably under the constraint of the remaining available power capacity of the current power distribution branch. The difference between this and the rated power is the power compression caused by other concurrent processes occupying the power distribution capacity. The productivity per unit time comes from the historical productivity field carried by the corresponding candidate replacement process process node in the BIM process library. This value is measured in terms of the amount of work completed by the corresponding machinery per unit time under the rated power, and the unit is consistent with the unit of measurement of the remaining work of the problem process. The calculation of the revised duration introduces the product of the concurrent power interference coefficient and the unit-time productivity as the denominator. Its physical meaning is the actual productivity of the candidate alternative process under the current limited power. A lower actual productivity than the rated productivity leads to an extended time required to complete the same remaining workload. The linear assumption is based on the approximate linear relationship between the productivity and power of commonly used machinery in substation construction within the range of 60% to 100% of the rated power. Below 60%, the machinery enters an uneconomical operating condition. At this point, the concurrent power interference coefficient of the candidate alternative process is usually below 0.6, and the corresponding net carbon gain is likely negative, thus it is naturally excluded in step S4. The remaining workload of the problematic process is calculated by subtracting the cumulative value of the completed workload up to the current collection period from the planned workload field value of the corresponding problematic process node in the process node list. The completed workload is recorded in real-time by the workload progress field of the problematic process node in the real-time mirror of the concurrent status.

[0032] The calculation boundary for predicted carbon emissions only includes carbon emissions from mechanical energy consumption, excluding carbon emissions from material consumption. This is because the target components completed by the candidate replacement process and the problem process are the same, and the types and quantities of building materials used are consistent. Material carbon emissions cancel each other out and do not affect the calculation result of net carbon gain. Carbon emissions from electric machinery are obtained by multiplying the corrected actual power, the regional power grid carbon emission factor, and the correction duration. Carbon emissions from fuel-powered machinery are obtained by multiplying the rated fuel consumption by the concurrent power interference coefficient to obtain the corrected fuel consumption, and then multiplying it by the diesel carbon emission factor and the correction duration. The corrected fuel consumption for fuel-powered machinery also incorporates the concurrent power interference coefficient because the actual fuel consumption of fuel-powered machinery under power-limited operating conditions is approximately linearly related to the power compression ratio. When a candidate replacement process includes both electric and fuel-powered machinery, the two types of carbon emissions are directly added to obtain the total predicted carbon emissions. This total value, in kilograms of CO2 equivalent, is stored in the predicted carbon emissions field of the corresponding candidate replacement process node in the real-time mirror of the concurrent state carrying the machinery's movement distance, for direct reading in the calculation of net carbon gain.

[0033] In one specific embodiment, step S4 includes: Multiply the remaining amount of work for the problematic process by the dynamic carbon emission intensity factor to obtain the carbon emissions required for the continued execution of the problematic process; Multiply the mechanical movement distance of the machinery corresponding to the candidate replacement process by the fuel consumption rate and the diesel carbon emission factor, add the fixed carbon emission amount of the machinery shutdown and start-up and the carbon emission quota value of the preparation operation standard of the candidate replacement process in the BIM process library, and obtain the carbon emission amount of the process replacement conversion. The net carbon gain is obtained by subtracting the sum of the predicted carbon emissions and the carbon emissions from the process replacement from the carbon emissions from the continued execution of the problematic process. The maximum net carbon gain in the candidate replacement process set is compared with zero. When the maximum net carbon gain is greater than zero, the node of the problematic process is replaced with the corresponding candidate replacement process in the BIM process association model, and a scheduling instruction containing the machine stop command, machine dispatch route and replacement process start timestamp is issued to the construction team's mobile terminal. When the maximum net carbon gain is not greater than zero, the remaining available power capacity of the current distribution branch, the spatial coordinates of each third-level process unit, and the concurrent power interference coefficient of each candidate replacement process in the real-time mirror of the concurrent state are used to form the current concurrent state feature vector. The current concurrent state feature vector is written into the concurrent applicability label of the candidate replacement process, and the difference between the dynamic carbon emission intensity factor and the upper limit of the carbon emission benchmark fingerprint bandwidth is superimposed on the upper limit of the carbon emission benchmark fingerprint bandwidth to obtain the updated upper limit of the carbon emission benchmark fingerprint bandwidth.

[0034] Specifically, in calculating the carbon emissions of the continued execution of the problematic process, the dynamic carbon emission intensity factor is taken as the arithmetic mean of the dynamic carbon emission intensity factors in each of the ten consecutive collection periods that triggered the problematic process, rather than a single period value. The basis for taking the mean is that the dynamic carbon emission intensity factor in each of the ten consecutive periods fluctuates slightly. The mean represents the stable carbon emission level of the problematic process under the current concurrent state, avoiding overestimation or underestimation of carbon emissions due to extreme values ​​in a single period. The fixed carbon emission amount for mechanical shutdown and restart is taken as a fixed empirical value of 0.5 kg CO2 equivalent per cycle. This value comes from the statistical average of energy consumption during the shutdown and restart of commonly used induction motors and diesel engines in substation construction. Its impact on a single replacement decision is smaller in magnitude than that of mechanical movement carbon emissions, but cannot be ignored. It is included in the calculation to ensure the integrity of the carbon emission amount of the process replacement. The carbon emission quota value for preparation work is derived from the preparation work field carried by the candidate alternative work nodes in the BIM work library. This field is filled in according to the preparation work list of the corresponding work in the construction organization design when the work node is established in step S1, covering the total carbon emission quota value of preparation links such as temporary support erection and formwork installation. The machinery dispatch route is generated in the construction site plane of the BIM work association model according to the shortest path principle based on the coordinates of the problem work operation position in the real-time mirror of the concurrent state of the machinery movement distance and the current three-dimensional spatial coordinates of the machinery corresponding to the candidate alternative work. The coordinates of the route nodes are sent to the mobile terminals of the construction team along with the dispatch instructions.

[0035] The current concurrent state feature vector is composed of three types of values: the sequence of remaining available power capacity values ​​for all power distribution branches in the current update cycle, the sequence of three-dimensional spatial coordinate values ​​for all machinery corresponding to the third-level process units, and the sequence of concurrent power interference coefficient values ​​for each candidate process in the candidate alternative process set. All three types of values ​​are derived from the current cycle readings of the corresponding fields in the real-time concurrent state mirror carrying the machinery movement distance. The vector dimension dynamically changes with the number of power distribution branches on the construction site, the number of ongoing third-level process units, and the size of the candidate alternative process set. The concurrent applicability tag is a dynamic attribute structure attached to the process nodes of the candidate alternative processes in the BIM process library. It includes four fields: historical cumulative verification count, historical cumulative successful replacement count, historical average net carbon benefit, and the current concurrent state feature vector. When the maximum net carbon benefit is not greater than zero, the current concurrent state feature vector is appended to the corresponding field, the historical cumulative verification count is incremented by one, and the historical average net carbon benefit is updated using a weighted average method. The effective time window for the upper bound of the carbon emission benchmark fingerprint bandwidth after the update is five minutes from the time of writing, that is, the subsequent ten consecutive collection cycles. After the five-minute window ends, the upper bound of the carbon emission benchmark fingerprint bandwidth will be restored to the original value written in step S1. The reason is that the real-time mirror of the concurrent state is refreshed every 30 seconds. After five minutes, the on-site concurrent state has changed sufficiently. Continuing to use the temporarily adjusted upper bound of bandwidth will interfere with the normal determination of carbon emission offset under the new concurrent state.

[0036] In one specific embodiment, the carbon emissions from process replacement are obtained as follows: the mechanical movement distance of the machine corresponding to the candidate replacement process in the concurrent real-time mirror carrying the mechanical movement distance is multiplied sequentially by the unit distance fuel consumption rate and diesel carbon emission factor of the corresponding machine in the BIM process library to obtain the carbon emissions from mechanical movement; the carbon emissions from mechanical movement are added to the fixed carbon emissions from machine shutdown and startup and the carbon emission quota value of the preparation operation standard for the candidate replacement process in the BIM process library to obtain the carbon emissions from process replacement.

[0037] The concurrent applicability label is updated in the following way: The historical cumulative virtual verification count of the candidate alternative process and the historical average net carbon benefit are updated by weighted average. The updated historical average net carbon benefit is obtained by multiplying the current net carbon benefit by the historical cumulative verification count minus one, adding the current net carbon benefit, and dividing by the historical cumulative verification count. The updated historical average net carbon benefit is written into the concurrent applicability label. When the cumulative number of historical verifications of a candidate alternative procedure is not less than five and the ratio of the cumulative number of successful replacements to the cumulative number of historical verifications is less than one-fifth, the concurrent applicability tag of the candidate alternative procedure is marked as inapplicable, and the candidate alternative procedure is removed from the candidate alternative procedure set.

[0038] The updated upper bound of the carbon emission benchmark fingerprint bandwidth is obtained as follows: the difference between the dynamic carbon emission intensity factor and the upper bound of the carbon emission benchmark fingerprint bandwidth is used to obtain the current excess amount; the current excess amount is added to the upper bound of the carbon emission benchmark fingerprint bandwidth to obtain the updated upper bound of the carbon emission benchmark fingerprint bandwidth, and the updated upper bound of the carbon emission benchmark fingerprint bandwidth is written into the corresponding process node in the process node list for comparison of the dynamic carbon emission intensity factor within the next five minutes.

[0039] Specifically, the fuel consumption rate per unit distance is derived from the mechanical parameter field of the machinery calling the corresponding candidate replacement process in the BIM process library. This field is filled in according to the rated fuel consumption rate of this model of machinery under no-load driving conditions in the machinery equipment technical manual, with the unit being liters per meter. The process of the machinery moving from its current position to the operation position of the problem process is considered a no-load dispatch condition. Calculating the carbon emissions of movement using the no-load fuel consumption rate is consistent with the actual energy consumption status. The weighted update calculation logic for the historical net carbon benefit average is as follows: multiply the value obtained by subtracting one from the historical cumulative verification count by the original historical net carbon benefit average to obtain the historical cumulative net carbon benefit total; add the current net carbon benefit value and divide by the historical cumulative verification count to obtain the updated historical net carbon benefit average. This calculation is equivalent to taking the arithmetic mean of the net carbon benefit of all historical verification counts. An incremental update method is used to avoid storing all historical net carbon benefit records. The counting rule for the cumulative successful replacement count is that the count is incremented by one each time the maximum net carbon benefit is greater than zero and the corresponding candidate replacement process is actually written into the BIM process association model to complete the node replacement. Cases where virtual verification is completed but no real replacement is performed are not counted as successful replacement counts. The removal of a candidate alternative procedure from the candidate alternative procedure index after it is not marked as applicable is a phased removal. The effective scope is limited to the current substation project construction cycle. When crossing projects, the concurrent applicability label of the candidate alternative procedure in the BIM procedure library is restored to the initial state.

[0040] The current operation of overlaying the excess amount onto the upper bound of the carbon emission benchmark fingerprint bandwidth is a single overlay. That is, the updated upper bound is obtained by adding the difference between the average dynamic carbon emission intensity factor over ten consecutive collection cycles of the trigger marker and the original upper bound of the carbon emission benchmark fingerprint bandwidth. The difference is not cumulatively overlaid because multiple overlays would cause the bandwidth upper bound to continuously expand, leading to malfunctions in the judgment of problematic processes. When the updated carbon emission benchmark fingerprint bandwidth upper bound is written to the corresponding process node in the process node list, it replaces the original upper bound field value in an overwrite manner, and the writing timestamp is recorded synchronously. The five-minute effective window is calculated starting from the writing timestamp. After the window ends, the original carbon emission benchmark fingerprint bandwidth upper bound field value is restored to the initial value calculated based on the historical sample set in step S1. The restoration operation is triggered by the edge computing node checking the difference between the writing timestamp and the current time at the beginning of each 30-second update cycle. If the difference exceeds 300 seconds, restoration is performed to ensure that the bandwidth upper bound returns to the normal judgment range based on historical data after the concurrent interference subsides. The five-minute window is set based on the fact that the shortest scheduling response cycle for concurrent processes at the substation construction site is usually no less than five minutes. Within this time, the concurrent state will not change fundamentally. Temporarily relaxing the upper limit of the bandwidth can avoid repeatedly triggering virtual verification calculations that are destined to fail during the duration of the same concurrent state.

[0041] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing carbon emissions during green construction of substations based on BIM, characterized in that, The method includes: Step S1: Based on the BIM process association model of the substation project, the substation construction process is decomposed into several third-level process units to obtain a list of process nodes. Step S2: Based on the measured power and carbon emission factor of the machinery corresponding to each level 3 process unit, calculate the dynamic carbon emission intensity factor of each level 3 process unit, and mark the level 3 process units whose dynamic carbon emission intensity factor continuously exceeds the upper limit of the carbon emission benchmark fingerprint bandwidth as problem processes. Step S3: Inject the candidate replacement process of the problem process into the real-time concurrent state mirror carrying the current measured power, spatial coordinates and remaining available capacity of the distribution branch of each third-level process unit. Calculate the concurrent power interference coefficient based on the ratio of the rated power of the candidate replacement process to the remaining available capacity of the distribution branch. Substitute the concurrent power interference coefficient into the carbon emission prediction formula of the candidate replacement process to obtain the predicted carbon emission of the candidate replacement process under the current concurrent state. Step S4: Calculate the net carbon gain based on the remaining workload of the problematic process, the dynamic carbon emission intensity factor, the predicted carbon emissions, and the carbon emissions from process replacement. Using the net carbon gain being greater than zero as a criterion, replace the node of the problematic process in the BIM process association model with the candidate alternative process, or write the current concurrent state feature vector into the concurrent applicability label of the candidate alternative process and update the upper bound of the carbon emission benchmark fingerprint bandwidth.

2. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 1, characterized in that, Step S1 includes: Based on the substation engineering design drawings, a three-dimensional information model containing five categories of professional components—civil structure, primary power equipment, secondary power equipment, cable lines, and temporary facilities—is established in BIM modeling software. Each component is assigned a component code, material type, design quantity, process number, and planned start and end time nodes for construction, thus obtaining a BIM process association model. Based on the process number of each component in the BIM process association model, the substation construction process is decomposed into first-level process, second-level process and third-level process in sequence. The third-level process is the smallest process unit. For each third-level process unit, process nodes carrying the type of machinery to be called, the rated power of the machinery and the planned amount of work are created to obtain a process node list. Based on the calling machine type of each process node in the process node list, candidate alternative processes that meet the construction result equivalence constraints, schedule constraints and resource availability constraints are retrieved from the BIM process library. A candidate alternative process index is established for the process nodes of each third-level process unit. The candidate alternative process index is written into the corresponding process node in the process node list to obtain an updated process node list carrying the candidate alternative process index. Based on the historical sample set of carbon emission intensity factors for each third-level process unit in the historical construction data of similar substations, the mean and standard deviation of the historical sample set of carbon emission intensity factors for each third-level process unit are calculated. The lower bound is the mean minus twice the standard deviation, and the upper bound is the mean plus twice the standard deviation. The carbon emission benchmark fingerprint bandwidth for each third-level process unit is obtained, and the carbon emission benchmark fingerprint bandwidth is written into the corresponding process node in the updated process node list.

3. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 1, characterized in that, Step S2 includes: With a data acquisition cycle of 30 seconds, the measured current and line voltage of the power supply lines of the electric machinery corresponding to each third-level process unit are multiplied, and the measured fuel flow rate and diesel calorific value of the fuel machinery outlet pipe are multiplied to obtain the cycle-average measured power of the machinery corresponding to each third-level process unit. Multiply the periodic average measured power by the regional power grid carbon emission factor or diesel carbon emission factor, and then divide by the amount of work completed by the corresponding third-level process unit within the collection period to obtain the dynamic carbon emission intensity factor of each third-level process unit. The dynamic carbon emission intensity factor is compared with the upper bound of the carbon emission benchmark fingerprint bandwidth carried by the corresponding process node in the process node list. The third-level process unit whose dynamic carbon emission intensity factor exceeds the upper bound of the carbon emission benchmark fingerprint bandwidth in ten consecutive collection cycles is marked as a problem process, and the problem process marking result is obtained. Based on the problem process marking results, from the candidate alternative process index of the corresponding process node in the process node list, candidate alternative processes that are currently unoccupied and whose planned completion time does not exceed 120% of the remaining planned time of the problem process are selected, thus obtaining a set of candidate alternative processes.

4. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 3, characterized in that, In step S3, the candidate replacement process for the problematic process is injected into the concurrent real-time mirror image carrying the current measured power, spatial coordinates, and remaining available capacitance of each third-level process unit, including: With an update cycle of 30 seconds, the average measured power of the machinery corresponding to each level-three process unit, the three-dimensional spatial coordinates of the machinery obtained by the ultra-wideband positioning tag, and the remaining available power capacity collected by the smart meters of each power distribution branch at the construction site are written into the corresponding process node of the BIM process association model with the process node number as the index key to obtain a real-time mirror of the concurrent state. Based on the process node number of each candidate alternative process in the candidate alternative process set, the remaining available power capacity of each power distribution branch within the corresponding work space of the candidate alternative process is read from the concurrent real-time image. The Euclidean distance between the mechanical three-dimensional space coordinates of the candidate alternative process and the work position coordinates of the problem process is written into the corresponding process node in the concurrent real-time image, thus obtaining the concurrent real-time image carrying the mechanical movement distance.

5. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 4, characterized in that, In step S3, the concurrent power interference coefficient is calculated based on the ratio of the rated power of the candidate replacement process to the remaining available capacitance of the distribution branch, including: Based on the real-time mirror image of the concurrent state of the carrying mechanical movement distance, the rated power of the candidate replacement process is compared with the remaining available power capacity of the power distribution branch within the corresponding work space of the candidate replacement process. When the rated power is not greater than the remaining available power capacity, the concurrent power interference coefficient is recorded as one; when the rated power is greater than the remaining available power capacity, the ratio of the remaining available power capacity to the rated power is recorded as the concurrent power interference coefficient, thus obtaining the concurrent power interference coefficient of each candidate replacement process.

6. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 5, characterized in that, In step S3, the concurrent power interference coefficient is substituted into the carbon emission prediction formula of the candidate replacement process to obtain the predicted carbon emission of the candidate replacement process under the current concurrent state, including: Multiply the concurrent power interference coefficient by the rated power of the candidate replacement process to obtain the corrected actual power of the candidate replacement process under the current concurrent state; Divide the remaining workload of the problematic process by the product of the unit time productivity of the candidate alternative process under rated power and the concurrent power interference coefficient to obtain the correction duration of the candidate alternative process under the current concurrent state. Multiply the corrected actual power by the regional power grid carbon emission factor and the correction duration, and add the product of the rated fuel consumption of the fuel machinery of the candidate replacement process, the concurrent power interference coefficient, the diesel carbon emission factor, and the correction duration to obtain the predicted carbon emissions of the candidate replacement process under the current concurrent state.

7. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 6, characterized in that, Step S4 includes: Multiply the remaining amount of work in the problematic process by the dynamic carbon emission intensity factor to obtain the carbon emissions required for the continued execution of the problematic process; Multiply the mechanical movement distance of the machinery corresponding to the candidate replacement process by the fuel consumption rate and the diesel carbon emission factor, add the fixed carbon emission amount of the machinery shutdown and start-up, and the carbon emission quota value of the preparation operation standard of the candidate replacement process in the BIM process library to obtain the carbon emission amount of the process replacement conversion. The carbon emissions from continuing the problematic process are subtracted from the sum of the predicted carbon emissions and the carbon emissions from the process replacement to obtain the net carbon gain. The maximum net carbon gain in the candidate replacement process set is compared with zero. When the maximum net carbon gain is greater than zero, the node of the problematic process is replaced with the corresponding candidate replacement process in the BIM process association model, and a scheduling instruction containing a machine stop command, machine dispatch route and replacement process start timestamp is issued to the construction team's mobile terminal. When the maximum net carbon gain is not greater than zero, the remaining available power capacity of the current distribution branch, the spatial coordinates of each third-level process unit, and the concurrent power interference coefficient of each candidate replacement process in the real-time mirror of the concurrent state are used to form the current concurrent state feature vector. The current concurrent state feature vector is written into the concurrent applicability label of the candidate replacement process, and the difference between the dynamic carbon emission intensity factor and the upper bound of the carbon emission benchmark fingerprint bandwidth is superimposed on the upper bound of the carbon emission benchmark fingerprint bandwidth to obtain the updated upper bound of the carbon emission benchmark fingerprint bandwidth.

8. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 7, characterized in that, The carbon emission reduction in the aforementioned process is obtained through the following methods: The mechanical movement distance of the candidate replacement process corresponding to the machine in the concurrent state real-time mirror carrying the mechanical movement distance is multiplied by the unit distance fuel consumption rate and diesel carbon emission factor of the corresponding machine in the BIM process library in sequence to obtain the mechanical movement carbon emission amount. The carbon emissions from mechanical movement, the carbon emissions from mechanical shutdown and startup, and the carbon emission quota for the preparation of candidate alternative processes in the BIM process library are added together to obtain the carbon emissions from the process replacement and conversion.

9. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 7, characterized in that, The concurrency suitability label is updated in the following ways: The historical cumulative virtual verification count and historical net carbon gain average of the candidate alternative process are updated in a weighted average manner. The updated historical net carbon gain average is obtained by multiplying the current net carbon gain by the historical cumulative verification count minus one, adding the current net carbon gain, and dividing by the historical cumulative verification count. The updated historical net carbon gain average is written into the concurrency applicability tag. When the cumulative number of historical verifications of the candidate alternative procedure is not less than five and the ratio of the cumulative number of successful replacements to the cumulative number of historical verifications is less than one-fifth, the concurrent applicability tag of the candidate alternative procedure is marked as inapplicable, and the candidate alternative procedure is removed from the candidate alternative procedure set.

10. The method for optimizing carbon emissions during green construction of substations based on BIM according to claim 7, characterized in that, The updated upper bound of the carbon emission benchmark fingerprint bandwidth is obtained in the following way: The current excess amount is obtained by subtracting the dynamic carbon emission intensity factor from the upper bound of the carbon emission baseline fingerprint bandwidth. The current excess is superimposed on the upper bound of the carbon emission benchmark fingerprint bandwidth to obtain the updated upper bound of the carbon emission benchmark fingerprint bandwidth. The updated upper bound of the carbon emission benchmark fingerprint bandwidth is then written into the corresponding process node in the process node list for dynamic carbon emission intensity factor comparison within the next five minutes.

Images