Commodity cross-domain circulation carbon emission accounting method and device based on multi-source big data fusion

Abstract

The application discloses a commodity cross-domain circulation carbon emission accounting method and device based on multi-source big data fusion, belongs to the technical field of carbon emission accounting and data processing, and solves the problem that carbon emission accounting is affected by the limitation of a single data source. The method comprises the following steps: acquiring multi-source heterogeneous big data during cross-domain circulation of a target product; decomposing the cross-domain circulation of the target product into multiple carbon emission circulation links; constructing a causal diagram according to a physical process of the cross-domain circulation of the commodity; dynamically determining weight weight and volume weight according to real-time weight utilization rate and volume utilization rate of a transportation tool in an intercity trunk line transportation link and an end distribution link; calculating a two-factor comprehensive allocation coefficient based on the weight weight and the volume weight; taking the two-factor comprehensive allocation coefficient as an intermediate node in the causal diagram, establishing a causal conduction path containing the two-factor comprehensive allocation coefficient, and calculating carbon emission based on the fused causal diagram. The application realizes accurate accounting of carbon emission.

Description

Technical Field

[0001] This invention belongs to the field of carbon emission accounting and data processing technology, specifically relating to a method and apparatus for carbon emission accounting of cross-regional commodity circulation based on multi-source big data fusion. Background Technology

[0002] Currently, with the continuous expansion of inter-regional commodity circulation, carbon emission accounting for the entire process of goods from warehousing and transportation to last-mile delivery has become a technological hotspot in the field of environmental management. E-commerce platforms, logistics systems, and warehousing systems generate massive amounts of order data, trajectory data, and energy consumption data every day. These data are large in volume, diverse in format, and frequently updated, exhibiting typical big data characteristics. This provides a big data foundation for refined carbon emission accounting, but also brings challenges to data fusion and processing.

[0003] Existing carbon emission accounting methods are mainly based on process analysis using a single data source. For example, they may rely solely on logistics companies' fuel consumption bills to calculate emissions during transportation, or solely on electricity consumption at warehousing facilities to calculate emissions during warehousing, or solely on the purchase volume of packaging materials to calculate emissions during packaging. While these methods can achieve relatively accurate quantification at specific stages, the data from each stage are independent and lack correlation, making it difficult to form a closed-loop accounting system covering the entire chain from goods leaving the warehouse to final delivery. This can easily lead to double counting or omission of emissions between stages.

[0004] Among the aforementioned related technologies, existing methods do not fully cover the entire process of commodity circulation. They often only conduct independent accounting for the transportation or packaging process, lacking systematic modeling of the entire chain of "warehousing-transportation-last-mile delivery". The carbon emission correlation between each link (such as the mutual influence between warehousing time and transportation batches) has not been effectively incorporated into the accounting system. Summary of the Invention

[0005] This application provides a method and apparatus for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion, which solves the problem in the prior art that the calculation of carbon emissions is affected by the limitation of a single data source.

[0006] On the one hand, embodiments of this application provide a method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion, the method including: Acquire multi-source heterogeneous big data during the cross-regional circulation of target products; The cross-regional circulation of the target product is broken down into multiple carbon emission circulation links, including packaging, warehousing, intercity trunk transportation, and last-mile delivery. Carbon emissions at each stage are defined as endogenous variables, and exogenous factors affecting carbon emissions at each stage are defined as exogenous variables. A causal graph containing directed edges of causal transmission between endogenous and exogenous variables is constructed based on the physical process of cross-regional commodity circulation. The causal graph contains at least one directed edge of causal transmission from carbon emissions at the warehousing stage to carbon emissions at the intercity trunk transportation stage, in order to characterize the causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. In the intercity trunk line transportation and last-mile delivery links, weight weight and volume weight are dynamically determined based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle. A two-factor comprehensive allocation coefficient is calculated for each package based on the weight weight and volume weight. Using the dual-factor comprehensive allocation coefficient as an intermediate node in the causal graph, a causal transmission path is established from the packaging stage variable through the package weight and package volume to the dual-factor comprehensive allocation coefficient and then to the carbon emissions in the transportation stage. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated. The fused causal graph includes the causal transmission path.

[0007] Optionally, acquiring multi-source heterogeneous big data during the cross-regional circulation of the target product includes: Obtain a product attribute dataset from an e-commerce platform, wherein the product attribute dataset includes at least product category, product sales volume, package volume, package weight, and shipping address; Obtain a packaging material dataset, which includes at least the type of packaging material used in the package, the weight of the packaging material, and the emission factor of the corresponding packaging material. Obtain a warehouse dataset, which includes at least the dwell time of goods in warehouses, distribution centers or delivery centers, the volume occupied by packages, the lighting power per unit volume and the power emission factor; Obtain a transportation vehicle parameter dataset, which includes at least the energy consumption per unit distance, energy density, rated cargo weight, rated cargo volume, and corresponding energy emission factor for each transportation vehicle. Obtain a transportation distance dataset, which includes at least the distance from the originating express station to the originating provincial transfer center, the distance from the originating provincial transfer center to the destination provincial transfer center, the distance from the destination provincial transfer center to the destination express station, and the last-mile delivery distance; Obtain a regional parameter dataset, which includes at least the regional electricity emission factors between different provinces or cities; Data from different data sources is standardized and converted according to a unified unit of measurement to obtain multi-source heterogeneous big data, in which volume is unified to cubic meters, weight to kilograms, distance to kilometers, energy consumption to liters per kilometer or kilowatt-hours per kilometer, and emissions to kilograms of carbon dioxide equivalent.

[0008] Optionally, the construction of a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation includes: Based on the physical process of cross-regional commodity circulation, the following causal transmission directed edges are constructed in the causal graph: The directed edge from the variable of packaging specification selection to the carbon emissions of the packaging process represents the direct causal impact of packaging material type and usage on carbon emissions of the packaging process. By selecting the variable of packaging specifications and pointing to the directed edges of package volume and weight to the carbon emissions of warehousing and intercity trunk transportation, we can characterize the indirect causal impact of packaging specifications on the space occupancy of warehousing and the allocation coefficient of transportation by changing package volume and weight. The directed edge from the warehousing strategy variable to the warehousing dwell time variable and then to the carbon emissions in the warehousing process represents the direct causal impact of warehousing strategy on carbon emissions in the warehousing process by influencing the parcel dwell time. The directed edge from carbon emissions in the warehousing stage to carbon emissions in the intercity trunk transportation stage represents the cross-stage causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and loading rate, thereby leading to changes in transportation emissions. Directed edges pointing from weather and road condition variables to carbon emissions in intercity trunk transportation represent the direct causal impact of external environmental factors on transportation efficiency and energy consumption. Directed edges pointing to the total carbon emissions from packaging, warehousing, intercity trunk transportation, and last-mile delivery represent the causal convergence relationship where emissions from each stage are aggregated into the total emissions.

[0009] Optionally, after constructing a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation, the method further includes: For each endogenous variable, a structural equation is established with its parent variable as the independent variable. The structural equation is used to quantify the magnitude of the causal effect of each parent variable on the endogenous variable. The structural equations include at least a structural equation for carbon emissions in the intercity trunk transportation segment. The independent variables of the structural equations include at least weather condition variables, road condition variables, carbon emissions in the warehousing segment, and packaging specification variables, in order to quantify the causal influence of the warehousing segment on the transportation segment through the causal transmission directed edge, as well as the causal influence of external environmental factors on the transportation segment.

[0010] Optionally, in the intercity trunk transportation and last-mile delivery stages, the weight weight and volume weight are dynamically determined based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle, including: Obtain the total weight of all packages currently loaded on the transport vehicle, and calculate the ratio of the total weight to the rated cargo capacity of the transport vehicle as the weight utilization rate of the current load; Obtain the total volume of all packages currently loaded on the transport vehicle, and calculate the ratio of the total volume to the rated cargo volume of the transport vehicle as the current volume utilization rate; The weight utilization rate is defined as the proportion of the weight utilization rate in the sum of the weight utilization rate and the volume utilization rate as the weight weight, and the volume utilization rate is defined as the proportion of the volume utilization rate in the sum of the weight utilization rate and the volume utilization rate as the volume weight.

[0011] Optionally, the calculation of the two-factor comprehensive allocation coefficient for each package based on the weight weight and volume weight includes: Obtain the total weight and total volume of all packages in the current transportation batch, and calculate the weight of each package as the proportion of the total weight and the volume of each package as the proportion of the total volume. Multiply the weight weight by the weight percentage of the package to obtain the weight distribution of the package; Multiply the volume weight by the volume percentage of the package to obtain the volume allocation component of the package; The weight allocation component and the volume allocation component are added together, and the sum is used as the two-factor comprehensive allocation coefficient for the package in the transportation batch.

[0012] Optionally, after constructing a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation, the method further includes: The exogenous variable is set to a first value through intervention, while keeping other exogenous variables unchanged. The values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the first expected value of the target variable under the first value. The exogenous variable is set to a second value through intervention, while keeping other exogenous variables unchanged. The values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the second expected value of the target variable under the second value. The difference between the first expected value under the first value and the second expected value under the second value is calculated as the total causal effect of the exogenous variable on the target variable, wherein the total causal effect includes direct causal effects and indirect causal effects: The direct causal effect is the causal influence of the exogenous variable directly pointing to the target variable through the directed edge in the causal graph. It is obtained by calculating the intervention effect after blocking all other paths except the direct directed edge in the causal graph. The indirect causal effect is the causal influence of the exogenous variable transmitted to the target variable through a directed path via at least one intermediate variable in the causal graph, and is obtained by subtracting the direct causal effect from the total causal effect.

[0013] Optionally, after calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: When the value of any exogenous variable changes, trace the complete path of that change propagating node by node along each causal directed edge in the causal graph; Record the increment of causal effect produced as the change propagates along each directed edge of causal transmission; The incremental causal effects are aggregated according to their respective causal transmission paths, and the effect proportion of each causal transmission path is output to identify the dominant causal transmission path and key intermediate variables that affect the total carbon emissions from this exogenous variable.

[0014] Optionally, after calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: Obtain actual observation data of the target order, and record the actual values ​​of each exogenous variable and the actual calculated values ​​of each endogenous variable; The value of the target exogenous variable is changed from the actual observed value to a preset counterfactual value, while keeping other exogenous variables unchanged; Following the topological order of the causal graph, starting from the exogenous variables, the values ​​of all endogenous variables are recalculated node by node to obtain the carbon emissions of each stage and the total carbon emissions under the counterfactual scenario. By comparing the carbon emissions at each stage under a counterfactual scenario with the actual calculated values, the causal effect of the change in this exogenous variable on the total carbon emissions can be quantified.

[0015] On the other hand, embodiments of this application provide a carbon emission accounting device for cross-regional commodity circulation based on multi-source big data fusion, the device comprising: The acquisition module is used to acquire multi-source heterogeneous big data when the target product is distributed across regions; The decomposition module is used to decompose the cross-regional circulation of the target product into multiple carbon emission circulation links, including packaging, warehousing, intercity trunk transportation, and last-mile delivery. The module is used to define the carbon emissions of each stage as endogenous variables and the exogenous factors affecting the carbon emissions of each stage as exogenous variables. Based on the physical process of cross-regional circulation of goods, a causal graph is constructed, which includes directed edges of causal transmission between endogenous and exogenous variables. The causal graph contains at least one directed edge of causal transmission from the carbon emissions of the warehousing stage to the carbon emissions of the intercity trunk transportation stage, so as to characterize the causal transmission relationship that changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. The determination module is used to dynamically determine weight weight and volume weight based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle in the intercity trunk transportation and last-mile delivery stages. The first calculation module is used to calculate the two-factor comprehensive allocation coefficient for each package based on the weight weight and volume weight. The second calculation module is used to establish a causal transmission path from the packaging stage variable through the package weight and package volume to the two-factor comprehensive allocation coefficient and then to the carbon emissions of the transportation stage, using the two-factor comprehensive allocation coefficient as an intermediate node in the causal graph. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated. The fused causal graph includes the causal transmission path.

[0016] In another aspect, embodiments of this application provide an electronic device, the device comprising: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the carbon emission accounting method for cross-regional circulation of goods based on multi-source big data fusion as described in the first aspect.

[0017] Compared with existing technologies, the carbon emission accounting method and apparatus for cross-regional commodity circulation based on multi-source big data fusion provided in this application constructs a causal graph containing carbon emissions and influencing factors at each circulation stage by integrating heterogeneous big data from multiple sources. It also introduces a two-factor comprehensive allocation coefficient to more accurately calculate emissions during transportation. This achieves closed-loop accounting of carbon emissions across the entire cross-regional commodity circulation chain, effectively avoiding duplicate calculations or omissions of emissions between stages in traditional methods. Furthermore, it reveals the correlation between carbon emissions at each stage, accurately calculates the carbon emissions of each commodity, and provides data support and decision-making basis for refined carbon emission management. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings. Figure 1 This is a flowchart illustrating a method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion, provided in one embodiment of this application. Figure 2 This is a schematic diagram of the structure of a carbon emission accounting device for cross-regional circulation of goods based on multi-source big data fusion, provided in another embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in another embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] To facilitate understanding of the embodiments of this application, further explanation and description will be provided below with reference to the accompanying drawings and specific embodiments. These embodiments do not constitute a limitation on the embodiments of this application. In the drawings, the dimensions and relative dimensions of components may be exaggerated for clarity and / or descriptive purposes. When exemplary embodiments can be implemented differently, a specific process sequence may be performed in a different order than that described. For example, two consecutively described processes may be performed substantially simultaneously or in the reverse order of their description. Furthermore, the same reference numerals denote the same components.

[0021] The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, unless the context clearly indicates otherwise, the singular forms “a” and “the” are intended to include the plural forms as well. Furthermore, when the terms “comprising” and / or “including” and variations thereof are used in this specification, it indicates the presence of the stated features, integrals, steps, operations, parts, components, and / or groups thereof, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, parts, components, and / or groups thereof. It should also be noted that, as used herein, the terms “substantially,” “about,” and other similar terms are used as approximate terms rather than as terms of degree, thus explaining the inherent biases in measurements, calculated values, and / or provided values ​​that would be recognized by one of ordinary skill in the art.

[0022] To address the problems in existing technologies, this application provides a method and apparatus for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion. In this application, by fusing heterogeneous big data from multiple sources, a causal graph is constructed containing carbon emissions and influencing factors at each stage of the circulation process. A two-factor comprehensive allocation coefficient is introduced to more accurately calculate emissions during transportation. This achieves closed-loop accounting of carbon emissions across the entire cross-regional commodity circulation chain, effectively avoiding duplicate calculations or omissions of emissions between stages in traditional methods. It also reveals the correlation between carbon emissions at each stage, accurately calculating the carbon emissions of each commodity, and providing data support and decision-making basis for refined carbon emission management.

[0023] The following section first introduces the carbon emission accounting method for cross-regional circulation of goods based on multi-source big data fusion provided in the embodiments of this application.

[0024] Figure 1 This illustration shows a flowchart of a method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion, according to an embodiment of this application. Figure 1 As shown, the carbon emission accounting method for cross-regional commodity circulation based on multi-source big data fusion can include: S101: Obtain multi-source heterogeneous big data when the target product is distributed across regions.

[0025] In this embodiment, relevant data can be collected from different data sources through methods such as manual input, file import, or API interface integration. This data may include basic product information, logistics and transportation records, warehousing operation logs, and packaging material lists.

[0026] S102 breaks down the cross-regional circulation of the target product into multiple carbon emission circulation links.

[0027] In this embodiment of the application, the carbon emission circulation process includes packaging, warehousing, intercity trunk transportation, and last-mile delivery; the entire process from the manufacturer's delivery to the consumer's receipt can be clearly defined according to physical flow and functional division, with the start and end points and activities of each stage clearly defined.

[0028] S103 defines the carbon emissions of each stage as endogenous variables and the exogenous factors affecting carbon emissions of each stage as exogenous variables. Based on the physical process of cross-regional circulation of goods, a causal graph containing directed edges of causal transmission between endogenous and exogenous variables is constructed.

[0029] In this embodiment, carbon emissions from the packaging process can be defined as endogenous variables, while the type and amount of packaging materials can be defined as exogenous variables. A causal graph with directed edges is constructed through the causal transmission between the endogenous and exogenous variables. The causal graph contains at least one directed edge that transmits carbon emissions from the warehousing process to the intercity trunk transportation process, in order to characterize the causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. For example, the causal relationship between variables can be drawn based on expert experience, historical data analysis, or domain knowledge, and connected by directed edges.

[0030] S104, in the intercity trunk line transportation and last-mile delivery links, dynamically determine the weight weight and volume weight based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle.

[0031] In this embodiment of the application, the current load and cargo volume can be estimated manually or obtained through sensor data based on the actual loading of the transport vehicle, and the ratio of these values ​​to the rated load weight and rated cargo volume can be calculated to determine the weight weight and volume weight.

[0032] S105 calculates a two-factor comprehensive allocation coefficient for each package based on weight and volume weight.

[0033] In this embodiment of the application, the allocation coefficient of the package in the current transportation batch can be calculated by weighted averaging based on the weight and volume of each package, as well as the pre-set weight weight and volume weight.

[0034] S106 uses the two-factor comprehensive allocation coefficient as an intermediate node in the causal graph to establish a causal transmission path from the packaging variable through the package weight and package volume to the two-factor comprehensive allocation coefficient and then to the carbon emissions in the transportation stage. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated.

[0035] In some embodiments, the calculated two-factor comprehensive allocation coefficient can be added as a new variable node to the causal graph, establishing a causal connection between it and upstream package weight and volume, as well as downstream carbon emissions from the transportation process. The carbon emissions of each stage and the total carbon emissions are calculated based on the fused causal graph, which includes causal transmission paths. For example, causal inference methods such as structural equation modeling or Bayesian networks can be used to calculate the carbon emissions of each stage based on the topological structure of the causal graph and the quantitative relationships between variables, and then the total carbon emissions can be obtained by summing them up.

[0036] In this embodiment, a causal graph containing carbon emissions and influencing factors at each stage of the distribution process is constructed by integrating multi-source heterogeneous big data. A two-factor comprehensive allocation coefficient is introduced to more accurately calculate emissions during transportation. This achieves closed-loop accounting of carbon emissions across the entire cross-regional distribution chain of goods, effectively avoiding the duplication or omission of emissions between stages in traditional methods. It also reveals the correlation between carbon emissions at each stage, accurately calculating the carbon emissions of each commodity, and providing data support and decision-making basis for refined carbon emission management.

[0037] In some embodiments, S101 may include: Obtain a product attribute dataset from an e-commerce platform. The product attribute dataset should include at least the product category, product sales volume, package volume, package weight, and shipping address. Obtain a packaging material dataset, which should include at least the type of packaging material used in the package, the weight of the packaging material, and the emission factor of the corresponding packaging material. Obtain a warehouse dataset, which should include at least the dwell time of goods in warehouses, distribution centers or delivery centers, the volume occupied by packages, the lighting power per unit volume and the power emission factor; Obtain a transportation vehicle parameter dataset, which should include at least the energy consumption per unit distance, energy density, rated cargo weight, rated cargo volume, and corresponding energy emission factor for each transportation vehicle. Obtain a transportation distance dataset, which includes at least the distance from the originating express station to the originating provincial transfer center, the distance from the originating provincial transfer center to the destination provincial transfer center, the distance from the destination provincial transfer center to the destination express station, and the last-mile delivery distance. Obtain a regional parameter dataset, which should include at least the regional electricity emission factors between different provinces or cities; Data from different data sources is standardized and converted according to a unified unit of measurement to obtain multi-source heterogeneous big data, in which volume is unified to cubic meters, weight to kilograms, distance to kilometers, energy consumption to liters per kilometer or kilowatt-hours per kilometer, and emissions to kilograms of carbon dioxide equivalent.

[0038] In this embodiment, the product attribute dataset is obtained from an e-commerce platform and includes product category, product sales volume, and package volume V. i Package weight (G) i Key information such as the origin and destination address is required. For example, the volume and weight of a package directly affect the choice of packaging materials, the use of storage space, and the loading efficiency and fuel consumption of transportation vehicles. The origin address is used to determine the starting point and route of transportation.

[0039] Packaging materials dataset, including the type of packaging materials used in the package, such as corrugated cardboard boxes or plastic bags, and the weight W of the packaging materials. (i,p) Corresponding emission factors Unlike existing technologies, this invention does not use industry average packaging usage for estimation. Instead, it performs a detailed calculation based on the actual type and weight of packaging materials used in each package to eliminate systematic errors caused by averaging.

[0040] Warehouse dataset, including the dwell time t of goods in warehouses, distribution centers, or delivery centers. i The volume occupied by the package is V i Lighting power per unit volume P L and electricity emission factors This invention introduces two dimensions—parcel volume and dwell time—to differentiate emissions across various scenarios. For example, dwell time affects the energy consumption of warehousing equipment, parcel volume affects the utilization rate of warehousing space, while lighting power and electricity emission factors are directly related to the carbon emissions generated by the electricity consumption of warehousing facilities.

[0041] A dataset of transportation parameters, including energy consumption per unit distance for various modes of transportation such as trucks, railways, airplanes, and electric tricycles. (Units are L / km for gasoline vehicles and kWh / km for electric vehicles), Energy density (Required for fuel-powered vehicles, not for electric vehicles) Rated cargo weight (Units are kg or t) and corresponding energy emission factors These parameters can be categorized and stored according to the type of transport vehicle, load capacity, and energy type. Since emissions at the boundaries between different modes of transport are difficult to define (e.g., manual labor, forklifts, or other handling equipment), lack unified standards, and account for a smaller proportion compared to mainline transport emissions, this invention does not establish a separate accounting step for emissions at these boundaries to reduce accounting bias. Compared to existing technologies, this invention specifically introduces the "rated cargo weight" parameter to provide a data basis for subsequent weight proportion allocation.

[0042] The transportation distance dataset includes the distances between the originating express station, the originating provincial transfer center, the destination provincial transfer center, and the destination express station, as well as the types of transportation vehicles used in each stage of transportation.

[0043] The regional parameter dataset includes transportation distances between different provinces or cities and regional electricity emission factors.

[0044] For data collected from different sources, their units need to be standardized, with volume uniformly expressed as m. 3 Weight is uniformly expressed in kg, distance in km, energy consumption in L / km, kWh / km or equivalent units, and emissions in kgCO2eq.

[0045] It's worth noting that for scenarios where the actual weight of packaging materials is unavailable, a method for estimating packaging materials based on product category and volume can be used. This alternative solution establishes a mapping database between product volume and commonly used packaging specifications, and then uses the statistical average packaging weight of products in the same category for estimation. Even when data collection conditions are limited (such as when third-party logistics platforms cannot provide order-by-order packaging data), the feasibility of the calculation can still be guaranteed.

[0046] In actual transportation, system failures, data upload interruptions, network delays, or human error may cause data loss for certain nodes or stages in the logistics path. A historical data-based interpolation method can be used to complete this data. When a segment of transportation distance data is missing, historical orders with the same origin and destination and the same mode of transport are retrieved from the database. The transportation distance for the corresponding stage is extracted, and the average value is calculated. This average value is then used as an estimate of the missing distance. The advantage of this approach is that it utilizes real historical data for completion, ensuring the estimation results are statistically representative.

[0047] When inconsistencies arise in timestamps from different data sources during actual data collection, a primary key alignment method based on order numbers is adopted. Using the order number from the e-commerce platform as a unique identifier, data from different data sources are associated according to the order number, ignoring timestamp differences between data sources. The advantages of this method are: the order number is a universally applicable and unique identifier across systems, eliminating the need for time calibration, and it is simple to implement with minimal computational cost.

[0048] In some other embodiments, S103 may include: Based on the physical process of cross-regional commodity circulation, the following causal transmission directed edges are constructed in the causal graph: The directed edge from the variable of packaging specification selection to the carbon emissions of the packaging process represents the direct causal impact of packaging material type and usage on carbon emissions of the packaging process. By selecting the variable of packaging specifications and pointing to the directed edges of package volume and weight to the carbon emissions of warehousing and intercity trunk transportation, we can characterize the indirect causal impact of packaging specifications on the space occupancy of warehousing and the allocation coefficient of transportation by changing package volume and weight. The directed edge from the warehousing strategy variable to the warehousing dwell time variable and then to the carbon emissions in the warehousing process represents the direct causal impact of warehousing strategy on carbon emissions in the warehousing process by influencing the parcel dwell time. The directed edge from carbon emissions in the warehousing stage to carbon emissions in the intercity trunk transportation stage represents the cross-stage causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and loading rate, thereby leading to changes in transportation emissions. Directed edges pointing from weather and road condition variables to carbon emissions in intercity trunk transportation represent the direct causal impact of external environmental factors on transportation efficiency and energy consumption. Directed edges pointing to the total carbon emissions from packaging, warehousing, intercity trunk transportation, and last-mile delivery represent the causal convergence relationship where emissions from each stage are aggregated into the total emissions.

[0049] In this embodiment, the directed edge pointing from the packaging specification selection variable to the carbon emissions of the packaging process serves to clarify the direct impact of packaging material type (e.g., cardboard boxes, plastic bags, cushioning materials, etc.) and their usage on the carbon emissions of the packaging process. Different materials have different carbon emission factors in their production and processing, and the greater the usage, the higher the carbon emissions of that stage are generally. For example, by selecting lightweight, recyclable, or biodegradable packaging materials and optimizing packaging design to reduce material usage, carbon emissions from the packaging process can be directly reduced.

[0050] This study aims to reveal the indirect impact of packaging specifications on subsequent logistics processes by tracing the carbon emissions from warehousing and intercity rail transport through the variables of packaging size selection, via parcel volume and weight. The choice of packaging size not only directly affects the packaging process but also influences space occupancy in warehousing and load distribution in intercity rail transport by altering the actual volume and weight of the parcel. For example, excessive or inappropriate packaging may increase parcel volume and weight, thus requiring more storage space and increasing energy consumption per unit area; during transportation, it may reduce vehicle loading rates or increase fuel consumption per trip, ultimately affecting carbon emissions in both warehousing and intercity rail transport.

[0051] A directed edge, pointing from warehousing strategy variables to warehousing dwell time variables and then to carbon emissions in the warehousing process, characterizes the impact of warehousing strategies (such as inventory management, order processing procedures, and sorting methods) on the dwell time of packages in warehouses, distribution centers, or delivery centers. The length of package dwell time directly relates to the energy consumption required for lighting, temperature control, and handling operations within warehousing facilities; the longer the dwell time, the higher the carbon emissions in the warehousing process. For example, implementing advanced inventory turnover strategies or optimizing warehouse layout can effectively shorten package dwell time, thereby reducing energy consumption and carbon emissions in the warehousing process.

[0052] The directed edge pointing from carbon emissions in the warehousing stage to carbon emissions in the intercity trunk transportation stage further refines the cross-stage impact of warehousing characteristics on carbon emissions in intercity trunk transportation. This directed edge represents how changes in warehousing dwell time affect the organization of transport batches and the loading rate of transport vehicles, thereby leading to changes in transportation emissions. For example, if parcels have a longer dwell time in the warehousing stage, it may create conditions for cargo consolidation and truck sharing, thereby increasing the loading rate of transport vehicles, reducing empty or underloaded transport, and thus reducing the carbon emissions per unit of cargo in the intercity trunk transportation stage. Conversely, if the dwell time is short, frequent departures may be necessary to meet timeliness requirements, resulting in a lower loading rate and increased transportation emissions.

[0053] Directed edges pointing to carbon emissions in intercity trunk transportation, derived from weather and road condition variables, are used to capture the direct impact of external environmental factors on transportation efficiency and energy consumption. Weather conditions (such as wind speed, rainfall, and temperature) and road conditions (such as congestion, road gradient, and road surface condition) are important exogenous variables affecting carbon emissions in intercity trunk transportation. Inclement weather or complex road conditions may lead to reduced vehicle speed, increased fuel consumption, and longer mileage, thereby directly increasing carbon emissions in intercity trunk transportation. For example, driving against the wind increases wind resistance, slippery roads increase rolling resistance, and congestion increases idling fuel consumption.

[0054] By pointing directed edges from carbon emissions in packaging, warehousing, intercity trunk transportation, and last-mile delivery to the total carbon emissions, it becomes clear that the carbon emissions at each stage of the distribution chain constitute the total carbon emissions of cross-regional commodity circulation. The total carbon emissions are the sum of emissions from each stage, reflecting the carbon footprint of the entire distribution chain. This convergence relationship forms the basis for overall carbon emissions accounting and analysis.

[0055] In this embodiment, by clarifying the direct and indirect impacts of packaging specification selection on packaging, warehousing, and transportation, as well as the impact of warehousing strategies on warehousing dwell time and thus on warehousing and transportation emissions, and combining this with the direct effects of external factors such as weather and road conditions on transportation efficiency, the constructed causal graph can more comprehensively and accurately reflect the real physical process and internal logic of carbon emissions from cross-regional commodity circulation.

[0056] In other embodiments, after constructing a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation, the method further includes: For each endogenous variable, a structural equation is established with its parent variable as the independent variable. The structural equation is used to quantify the magnitude of the causal effect of each parent variable on the endogenous variable. The structural equation model should include at least a structural equation for carbon emissions in the intercity trunk transportation segment. The independent variables of this structural equation model should include at least weather conditions, road conditions, carbon emissions in the warehousing segment, and packaging specifications, in order to quantify the causal influence of the warehousing segment on the transportation segment through causal transmission directed edges, as well as the causal influence of external environmental factors on the transportation segment.

[0057] In this embodiment, for each endogenous variable in the causal graph, a structural equation can be established with all its parent variables (i.e., variables that directly point to the endogenous variable) as independent variables. This equation is typically expressed as Y = f(X1,X2, ..., X...). n The form is ε), where Y represents the endogenous variable, X1 to X... n Let f represent the parent variable, f be the function describing the relationship, and ε be the error term. It should be noted that the exogenous variables in this application are selected based on the physical processes and domain knowledge of cross-domain commodity circulation, specifically including but not limited to: packaging specification selection variables (such as packaging material type and quantity), warehousing strategy variables (such as inventory turnover strategy), weather condition variables (such as temperature, wind speed, and rainfall), and road condition variables (such as road congestion). These variables are directly related to the physical characteristics of the circulation process and are not infinitely expandable. Those skilled in the art can determine the above exogenous variables based on the actual physical processes of commodity circulation.

[0058] By conducting statistical analysis on historical data, such as using the least squares method or maximum likelihood estimation, the parameters in the structural equation (e.g., regression coefficients) can be estimated. The magnitude and sign of these parameters directly reflect the strength and direction of the causal effect of each parent variable on the endogenous variable. For example, if a linear regression model is used, Y = β0 + β1 × X1 + β2 × X2 + ... + β n ×X n +ε, then the coefficient β i That is, the parent variable X is quantified. i The magnitude of the causal effect on the endogenous variable Y. The regression coefficient β mentioned above. i The method for determining the value is as follows: collect data containing the values ​​of each parent variable X. i Using historical data of the target variable Y, a linear regression model is fitted using the least squares method to minimize the sum of squared residuals, thereby estimating the specific values ​​of each regression coefficient.

[0059] To analyze and manage carbon emissions more precisely, this application specifically emphasizes establishing a structural equation model for carbon emissions in intercity trunk transportation. The independent variables of this structural equation model include at least weather condition variables, road condition variables, carbon emission variables in the warehousing stage, and packaging specification variables. Weather condition variables can encompass environmental factors such as temperature, humidity, wind speed, and rainfall, which directly affect the energy consumption and operating efficiency of transportation vehicles. Road condition variables can include road congestion levels, road surface type, and gradient, which also significantly impact transportation efficiency and fuel consumption. The carbon emission variable in the warehousing stage, as an independent variable, aims to quantify the cross-stage causal impact of warehousing operational strategies (such as warehousing dwell time and transportation batch organization methods) on carbon emissions in intercity trunk transportation by influencing transportation efficiency and loading rates. The packaging specification variable indirectly affects the allocation coefficient in the transportation stage by changing the volume and weight of packages, thus influencing transportation carbon emissions. By constructing and solving this structural equation, the specific impact of the aforementioned external environmental factors and upstream links on carbon emissions in intercity trunk transportation can be accurately quantified, providing data support for identifying key influencing factors and formulating targeted emission reduction measures.

[0060] Through the aforementioned technical solution, this application transforms the qualitative causal relationships in a causal diagram into a quantitative mathematical model, thereby enabling precise quantification of the causal effects of each parent variable on the endogenous variable. This overcomes the limitation of relying solely on causal diagrams to accurately quantify causal effects, elevating the analysis of carbon emission influencing factors from a qualitative to a quantitative level. Particularly in intercity trunk transportation, by constructing structural equations that include weather conditions, road conditions, carbon emissions from warehousing, and packaging specifications as independent variables, the specific impact of external environmental factors and upstream processes on transportation carbon emissions can be clearly revealed and quantified. This quantitative capability allows managers to gain a deeper understanding of the driving mechanisms of carbon emissions, identify key factors influencing carbon emissions and their degree of impact, thereby achieving more efficient and precise carbon emission management and optimization.

[0061] In some other embodiments, S104 may include: Obtain the total weight of all packages currently loaded on the transport vehicle, and calculate the ratio of the total weight to the rated cargo capacity of the transport vehicle as the weight utilization rate of the current load; Obtain the total volume of all packages currently loaded on the transport vehicle, and calculate the ratio of the total volume to the rated cargo volume of the transport vehicle as the current volume utilization rate; The proportion of weight utilization rate in the sum of weight utilization rate and volume utilization rate is determined as weight weight, and the proportion of volume utilization rate in the sum of weight utilization rate and volume utilization rate is determined as volume weight.

[0062] In this embodiment, the total weight of all packages on the current transport vehicle can be accurately obtained first through vehicle-mounted sensors, aggregated package weight data recorded by the logistics management system, or by weighing the loaded transport vehicle and subtracting its own weight. By calculating the ratio of the actual total weight to the rated load capacity, a value between 0 and 1 can be obtained. This value intuitively reflects the actual carrying efficiency and utilization of the transport vehicle in terms of weight. For example, if a truck with a rated load capacity of 10 tons actually carries 8 tons of goods, its weight utilization rate is 0.8. This weight utilization rate is a crucial foundational data for the subsequent dynamic determination of weight weights, ensuring the quantification of the weight dimension occupancy of transport resources.

[0063] Secondly, the total volume of all packages on the transport vehicle can be obtained by summing up the volume data of each package recorded by the logistics management system, or by measuring the loading space using technologies such as 3D scanning and image recognition. By calculating the ratio of the actual total volume to the rated cargo volume, a value between 0 and 1 can be obtained, reflecting the actual spatial utilization efficiency of the transport vehicle. For example, if a truck with a rated cargo volume of 60 cubic meters actually carries 45 cubic meters of goods, its volume utilization rate is 0.75. This volume utilization rate, together with the aforementioned weight utilization rate, forms the basis for the two-factor allocation, ensuring a comprehensive consideration of the multi-dimensional use of transport resources.

[0064] Finally, the proportion of weight utilization rate in the sum of weight utilization rate and volume utilization rate is determined as the weight weight, and the proportion of volume utilization rate in the sum of weight utilization rate and volume utilization rate is determined as the volume weight. In actual transportation, the load-bearing capacity and cargo space of transportation vehicles are often mutually restrictive. For example, for lightweight goods, the transportation vehicle may reach its volume limit first, while its weight is far from reaching it; for heavy goods, it may reach its weight limit first, while its volume still has room for maneuver. By normalizing the weight utilization rate and volume utilization rate, that is, calculating their respective proportions in the sum of the two, the degree to which weight and volume dominate the consumption of transportation resources in the current transportation batch can be dynamically reflected. For example, if the calculated weight utilization rate is 0.8 and the volume utilization rate is 0.2, then the sum of the two is 1.0. At this time, the weight weight is 0.8 / 1.0 = 0.8, and the volume weight is 0.2 / 1.0 = 0.2. This dynamic adjustment mechanism allows the weight allocation to flexibly adapt to different cargo types and loading conditions, avoiding the deviations that may be caused by using fixed weights, and providing a reliable basis for the subsequent accurate calculation of the two-factor comprehensive allocation coefficient for each package.

[0065] In this embodiment, the actual occupancy of transportation resources across different dimensions is quantified by accurately acquiring the real-time weight and volume utilization rates of transportation vehicles. Based on this, the weights are dynamically adjusted by using the proportion of each utilization rate in the sum of the two as weights, allowing for flexible adaptation to differences in the weight and volume characteristics of goods in different transportation batches. For example, when transportation vehicles are primarily constrained by weight, the weight weight is increased accordingly; when primarily constrained by volume, the volume weight is increased. This dynamic adjustment mechanism avoids the limitations of traditional fixed weight allocation, enabling a more accurate reflection of the actual contribution to transportation carbon emissions when calculating the dual-factor comprehensive allocation coefficient for each package. This not only improves the accuracy and fairness of carbon emission accounting but also provides data support for logistics companies to optimize loading strategies and improve transportation efficiency, thereby contributing to more effective carbon emission management and reduction targets.

[0066] In some other embodiments, S105 may include: Get the total weight and total volume of all packages in the current shipping batch, and calculate the weight percentage of each package relative to the total weight as the weight percentage of that package, and the volume percentage of each package relative to the total volume as the volume percentage of that package. Multiply the weight weight by the weight percentage of the package to obtain the weight allocation of the package. Multiply the volume weight by the volume percentage of the package to obtain the volume allocation component of the package; The weight allocation component and the volume allocation component are added together, and the sum is used as the two-factor comprehensive allocation coefficient for the package in the transportation batch.

[0067] In this embodiment, the weight data of all packages to be transported in the current transport batch is first acquired and summed using a logistics information system or real-time weighing / measuring equipment to obtain the total weight. Simultaneously, the volume data of all packages is acquired and summed to obtain the total volume. Then, the weight percentage of each package relative to the total weight is calculated as the package's weight percentage, and the volume percentage of each package relative to the total volume is calculated as the package's volume percentage. These percentage data quantify the relative contribution of each package to the overall weight and volume capacity of the current transport batch.

[0068] Based on this, the weight weight previously dynamically determined according to the real-time utilization rate of the transport vehicle is multiplied by the weight percentage of the package to obtain the weight allocation component of the package. This component reflects the share of carbon emissions that the package should bear due to its weight, under current transport conditions, taking into account the weight utilization rate of the transport vehicle and its impact on carbon emissions.

[0069] Similarly, the dynamically determined volume weight is multiplied by the volume percentage of the package to obtain the volume allocation component of the package. This component reflects the share of carbon emissions that the package should bear due to its volume, under current transportation conditions, taking into account the volume utilization rate of the transportation vehicle and its impact on carbon emissions.

[0070] Finally, the weight allocation component and the volume allocation component calculated above are added together, and the sum is used as the two-factor comprehensive allocation coefficient for the package in the transportation batch. This coefficient comprehensively considers the occupancy of the package's weight and volume on the transportation vehicle's capacity, and combines the relative importance of weight and volume to carbon emissions reflected by the real-time utilization rate of the transportation vehicle, thus providing a comprehensive and dynamic basis for carbon emission allocation.

[0071] Through the above technical solution, this application enables refined allocation of carbon emissions in the transportation process at the individual package level. This two-factor comprehensive allocation coefficient not only considers the physical properties of the package itself (weight and volume), but more importantly, it incorporates the weight and volume weights determined by the real-time utilization rate of the transportation vehicle. This allows the allocation results to dynamically adapt to actual transportation scenarios, making the carbon emission calculation results for each package more accurate and fair. This provides a solid data foundation for subsequent carbon emission management, optimization of transportation strategies, and the development of differentiated carbon footprint labels.

[0072] In some other embodiments, after S103, the method may further include: By intervening and setting the exogenous variable to its first value, while keeping other exogenous variables unchanged, the values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the first expected value of the target variable under the first value. By intervening and setting the exogenous variable to a second value, while keeping other exogenous variables unchanged, the values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the second expected value of the target variable under the second value. The difference between the first expected value under the first value and the second expected value under the second value is calculated as the total causal effect of the exogenous variable on the target variable. This total causal effect includes both direct and indirect causal effects. The direct causal effect is the causal influence of an exogenous variable through the directed edge in the causal graph that directly points from the exogenous variable to the target variable. It is obtained by calculating the intervention effect after blocking all other paths in the causal graph except for the direct directed edge. Indirect causal effect is the causal influence of an exogenous variable transmitted to the target variable through a directed path via at least one intermediate variable in a causal graph. It is obtained by subtracting the direct causal effect from the total causal effect.

[0073] In this embodiment, the method simulates the impact of exogenous variable changes on the target variable through an "intervention" operation. First, an exogenous variable of interest is selected and assigned a specific "first value," while keeping the values ​​of all other exogenous variables in the causal graph unchanged. Then, based on the constructed causal graph and its structural equation, the expected values ​​of all endogenous variables are recalculated one by one according to the topological order of the causal graph (i.e., from cause to effect) until the "first expected value" of the target variable (e.g., total carbon emissions or carbon emissions at a specific stage) under the "first value" is obtained. Similarly, the exogenous variable is assigned another "second value," and the same calculation process is repeated to obtain the "second expected value" of the target variable under the "second value." This intervention calculation simulates the systematic response caused by a change in a single exogenous variable while keeping other conditions constant. This can be achieved through the do-calculus operation in a structural causal model (SCM). In SCM, intervening in a variable (doing a function) means forcing its value to a constant, cutting off all incoming causal edges pointing to itself, and then making inferences based on the new model structure. Specifically, the structural equations in the causal graph can be used to substitute the values ​​of the intervention variable into the corresponding equations, and the expected values ​​of downstream variables can be calculated layer by layer along the causal path.

[0074] Subsequently, by comparing the differences in the expected values ​​of the target variable under the two different intervention values, the total effect caused by the change in the exogenous variable can be obtained. This is achieved by calculating the difference between the first expected value under the first intervention value and the second expected value under the second intervention value, which is taken as the total causal effect of the exogenous variable on the target variable. This difference represents the average change in the target variable when the exogenous variable changes from the "first value" to the "second value," reflecting the overall causal influence of the exogenous variable on the target variable.

[0075] To gain a deeper understanding of causal mechanisms, the total causal effect is further decomposed into direct causal effects and indirect causal effects. Direct causal effects refer to the influence of an exogenous variable acting directly on the target variable without any intermediate variables. To isolate this direct influence, the causal graph needs to be modified, i.e., "blocking all paths except for direct directed edges." This means that during computation, the path from the exogenous variable to the target variable through intermediate variables must be cut off, retaining only the edges directly pointing from the exogenous variable to the target variable. Then, on this modified causal graph, the intervention calculation is performed again, and the expected change in the target variable is obtained, which is the direct causal effect. In structural causal models, this can be achieved by modifying the model structure. For example, for exogenous variable X and target variable Y, if the path X->M->Y exists, to calculate the direct effect of X on Y, the influence of M on Y needs to be fixed, or the influence of X on M needs to be cut off, so that X can only influence Y through the direct edge X->Y. This typically involves fixing the value of the intermediate variable M to a baseline value (e.g., the expected value of M before the X intervention), or adjusting all structural equations along the path through M to make them unaffected by changes in X when calculating the effect of X on Y.

[0076] Finally, indirect causal effects refer to the effects produced by exogenous variables influencing one or more intermediate variables, which in turn further influence the target variable. Since the total causal effect is the sum of direct and indirect effects, once the total and direct causal effects are calculated, the indirect causal effect can be obtained by subtracting the direct causal effect from the total causal effect. This calculation method avoids the difficulty of directly modeling and calculating all complex indirect paths, simplifying the analysis process.

[0077] This application accurately obtains the "total causal effect" caused by changes in a specific exogenous variable by intervening to set different values ​​and recalculating the expected value of the target variable based on the topological order of the causal graph. More importantly, this application further decomposes the "total causal effect" into "direct causal effect" and "indirect causal effect." The "direct causal effect" is calculated by blocking all paths except direct directed edges, thus isolating the direct influence of the exogenous variable on the target variable. The "indirect causal effect" is obtained by subtracting the direct causal effect from the total causal effect, revealing the complex path of the exogenous variable's influence on the target variable through at least one intermediate variable (e.g., storage dwell time, transportation batch organization method, etc.), avoiding blind intervention and significantly improving the guiding value and decision support capability of carbon emission accounting results.

[0078] In other embodiments, after calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: When the value of any exogenous variable changes, trace the complete path of that change propagating node by node along each causal directed edge in the causal graph; Record the increment of causal effect produced as the change propagates along each directed edge of causal transmission; The incremental causal effects are aggregated according to their respective causal transmission paths, and the effect proportion of each causal transmission path is output to identify the dominant causal transmission path and key intermediate variables that affect the total carbon emissions of this exogenous variable.

[0079] In this embodiment, when the value of any exogenous variable changes, this change serves as an initial perturbation, propagating along directed edges in a pre-constructed causal graph. Starting from the changed exogenous variable, the system traverses all downstream nodes in the causal graph using a graph traversal algorithm (such as depth-first search or breadth-first search), identifying and recording all complete causal transmission paths originating from the exogenous variable and ultimately reaching the target variable (e.g., total carbon emissions).

[0080] During the path tracing process described above, as a change propagates along each causal transmission directed edge from a parent node to its child node, the system quantifies the specific impact of the parent node's change on the child node—the causal effect increment—based on a pre-established structural equation. For example, if a change in the parent node causes a specific magnitude of change in the child node, that change is the causal effect increment on that directed edge. By recording the increment on each directed edge, the hierarchical transmission process of causal effects in the graph can be meticulously depicted. After recording the causal effect increments on all causal transmission directed edges, the system accumulates or comprehensively calculates all causal effect increments belonging to the same complete causal transmission path, thereby obtaining the total contribution of that path to the target variable.

[0081] Subsequently, by comparing the contributions of different pathways, the proportion of each causal transmission pathway in the total effect can be calculated, i.e., the effect percentage. For example, if the effect percentage of a certain pathway is significantly higher than that of other pathways, it indicates that this pathway is the main way in which exogenous variables affect total carbon emissions; if an intermediate variable appears in multiple dominant pathways, it is a key intermediate variable.

[0082] Through the aforementioned technical solution, this application overcomes the limitations of merely quantifying the overall causal effect, and deeply reveals the specific mechanisms by which exogenous variables influence total carbon emissions. By tracing the complete path of change propagating node by node along each directed edge of the causal transmission graph, and recording the increment of the causal effect on each edge, this application can decompose complex causal relationships into understandable and specific transmission chains. Furthermore, by aggregating these increments according to their respective causal transmission paths and outputting their effect proportions, this application can clearly identify the dominant causal transmission paths and key intermediate variables affecting total carbon emissions. This enables policymakers not only to know "what" affects carbon emissions, but more importantly, to know "how" it affects them and "through which specific links" it affects them.

[0083] In other embodiments, after calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: Obtain actual observation data of the target order, and record the actual values ​​of each exogenous variable and the actual observed values ​​of each endogenous variable; The value of the target exogenous variable is changed from the actual observed value to a preset counterfactual value, while keeping other exogenous variables unchanged; Following the topological order of the causal graph, starting from the exogenous variables, the values ​​of all endogenous variables are recalculated node by node to obtain the carbon emissions of each stage and the total carbon emissions under the counterfactual scenario. By comparing the carbon emissions at each stage under the counterfactual scenario with the actual observed values, the causal effect of the change in this exogenous variable on the total carbon emissions can be quantified.

[0084] Specifically, the first step is to obtain the actual observation data of the target orders and record the actual values ​​of each exogenous variable and the actual observation values ​​of each endogenous variable.

[0085] Subsequently, the value of the target exogenous variable is modified from the actual observed value to a pre-defined counterfactual value, while keeping other exogenous variables unchanged. For example, for an order that actually used a certain type of plastic packaging, its packaging specification selection variable can be modified to the counterfactual value of using biodegradable cardboard boxes; or, for an order that actually stayed in the warehouse for a long time, its warehousing strategy variable can be modified to a faster turnover strategy, thereby setting a shorter storage stay time. When making such modifications, it is strictly ensured that all other exogenous variables (such as weather conditions, road conditions, etc.) besides the target exogenous variable remain at their original actual observed values ​​to ensure that the quantified causal effect is attributed only to changes in the target exogenous variable.

[0086] Based on this, following the topological order of the causal graph, the values ​​of all endogenous variables are recalculated node by node, starting with the exogenous variables, to obtain the carbon emissions at each stage and the total carbon emissions under the counterfactual scenario. For example, the causal relationships between variables revealed by the previously constructed causal graph, and the structural equations established for each endogenous variable (as described above), can be used as input, with the modified counterfactual exogenous variable values. Starting from these modified exogenous variables, the system updates the values ​​of the endogenous variables directly affected by them sequentially along the directed edges of the causal graph. These changes in endogenous variables will further propagate downstream along the causal chain, affecting subsequent endogenous variables, until all endogenous variables (including carbon emissions from packaging, warehousing, intercity trunk transportation, and last-mile delivery) and the final total carbon emissions are recalculated, thus obtaining the carbon emission results under the counterfactual scenario.

[0087] Finally, by comparing the carbon emissions at each stage under the counterfactual scenario with the actual observed values, the causal effect of the change in this exogenous variable on the total carbon emissions is quantified. By comparing the recalculated carbon emissions at each stage and the total carbon emissions under the counterfactual scenario with the actual observed carbon emissions at each stage and the total carbon emissions for the target order, the increase or decrease in carbon emissions due to the change from the actual value to the counterfactual value of the target exogenous variable can be accurately calculated. For example, if the actual total carbon emissions are X, and the total carbon emissions under the counterfactual scenario are Y, then YX represents the causal effect of the change in this exogenous variable on the total carbon emissions.

[0088] By simulating scenarios such as "how much carbon emissions would have been reduced if more environmentally friendly packaging materials had been chosen" or "how much carbon emissions could have been saved by optimizing warehouse management and shortening storage time," companies can quantify the emission reduction potential of different optimization measures, thereby developing more targeted and effective carbon management plans. This capability of "hypothetical analysis" based on real data significantly enhances the practicality and decision support value of carbon emission accounting results, making carbon emission reduction no longer a vague goal, but a quantifiable and optimizable concrete action.

[0089] The following example illustrates the distance calculation process. When calculating the carbon emissions at each stage and the total carbon emissions, a set of online e-commerce orders is obtained. This set includes at least one online shopping order, and each order corresponds to at least one product package. The entire process of cross-regional product distribution is broken down into four independent but data-linked stages: packaging, warehousing, intercity transportation, and last-mile delivery. For product orders... Its total carbon emissions Calculated by the following formula: ; in, Indicates package The total greenhouse gas emissions generated throughout the entire online shopping process, This indicates the carbon emissions from the packaging process. This indicates the carbon emissions during the warehousing process. This indicates the carbon emissions from intercity trunk line transportation. This indicates the carbon emissions during the last-mile delivery process.

[0090] By defining the system boundaries as described above, this invention avoids emissions omissions caused by estimating emissions only for a single transportation or packaging stage, and can identify and summarize different emission sources in the online shopping fulfillment process at each stage. Unlike existing methods that only focus on fuel consumption during transportation, this invention achieves full-chain coverage from packaging to last-mile delivery.

[0091] The accounting models for each sub-stage are as follows: Carbon emissions from packaging The packaging stage refers to the process of using packaging materials such as corrugated cardboard boxes, plastic bags, woven bags, and document envelopes to meet the requirements of transportation protection, warehousing and stacking, label identification, and last-mile delivery before goods leave the warehouse or are shipped. Emissions from the packaging stage mainly originate from the implicit carbon emissions associated with the production, processing, and use of packaging materials, falling under the category of upstream emissions. This invention does not use industry average packaging usage for estimation, but rather is based on parcels... The actual type and weight of packaging materials used are calculated using the following formula: ; in, Indicates goods The weight of Class P packaging materials is expressed in kg. This represents the emission factor of Class P packaging materials, expressed in kgCO2eq / kg.

[0092] Carbon emissions from warehousing Emissions in the warehousing process mainly originate from the electricity consumption for lighting corresponding to the space occupied by goods during their stay in warehouses, distribution centers, or delivery centers. Traditional methods typically allocate emissions simply based on the total storage area or total energy consumption, resulting in goods of different volumes and residence times being assigned the same unit emission intensity, leading to coarse-grained emissions. This invention focuses on incorporating package volume, storage residence time, unit volume lighting power, and electricity emission factors into the calculation, using the following formula: ; in, Indicates package The time spent in a warehouse, distribution center, or delivery center is measured in hours (h). Indicates package Volume, in m 3 ; This indicates the lighting power per unit volume of storage space, expressed in kW / m². 3 ; This represents the electricity emission factor, expressed in kgCO2eq / kWh.

[0093] By incorporating both storage dwell time and package volume into the calculation, this invention reflects the impact of different package volumes, storage cycles, and regional electricity emission factors on carbon emissions during warehousing. This model effectively distinguishes between the warehousing emissions of "large but fast-moving" and "small but long-stocked" goods, avoiding the misconception that "large items necessarily have high emissions." Compared to existing methods that allocate warehousing energy consumption solely based on building area, this invention introduces a two-dimensional "volume × time" calculation mechanism, significantly improving the fairness and accuracy of emission allocation during warehousing.

[0094] Carbon emissions from intercity trunk transportation The intercity trunk transportation segment is used to calculate emissions generated by parcels moving between cities or provincial logistics networks. This invention further divides intercity trunk transportation into three sub-stages: the first stage is the transportation stage from the originating express station to the originating provincial transfer center, denoted as the OS-OC stage; the second stage is the transportation stage from the originating provincial transfer center to the destination provincial transfer center, denoted as the OC-DC stage; and the third stage is the transportation stage from the destination provincial transfer center to the destination express station, denoted as the DC-DS stage. This segment is a key link in carbon emission accounting for cross-regional online shopping. Compared with existing technologies, this invention refines the transportation stage into three segments and calculates emissions separately, facilitating the identification of emission reduction potential at different stages. Therefore, the carbon emission calculation formula for the intercity trunk transportation segment is: ; in, This indicates the emissions during the stage from the originating express station to the originating provincial transfer center; This indicates the emissions during the stage from the originating provincial transshipment center to the destination provincial transshipment center; This indicates the emissions during the stage from the destination provincial transfer center to the destination express station.

[0095] For the OS-OC stage, the transportation vehicle is a truck, and the calculation formula is: ; in, Indicates package The transportation distance from the originating express station to the originating provincial transfer center is expressed in km. This indicates the diesel consumption of a truck per kilometer, expressed in L / km. This indicates the density of diesel fuel, expressed in kg / L. Indicates diesel emission factor; Indicates the rated load capacity of the truck; Indicates the weight of the package; This indicates the proportion of the package weight to the vehicle's load capacity, used for allocation, i.e., the two-factor comprehensive allocation coefficient in the OS-OC stage.

[0096] For the OC-DC phase, the means of transport can include road freight, air freight, and rail freight, calculated according to the mode of transport: ; in, This indicates the transportation distance from the provincial-level transit center at the origin to the provincial-level transit center at the destination, in km. This indicates the mode of transportation actually used in the current transportation phase. ; Indicates the mode of transport Energy consumption per unit distance; The density of energy (not used if electrical energy is used) is expressed in kg / L. Indicates the mode of transport Energy emission factors; Indicates the rated cargo capacity of the transport vehicle; Indicates the weight of the package; This represents the proportion of the package weight to the vehicle's load capacity, i.e., the two-factor comprehensive allocation coefficient in the OC-DC stage. This formula supports flexible combinations of road, rail, and air transport modes, and compared to previous technologies, this invention can adapt to complex multimodal transport scenarios.

[0097] For the DC-DS phase, where the means of transport is a truck, the calculation formula is: ; in, Indicates package The transportation distance from the originating express station to the originating provincial transfer center is expressed in km. This indicates the diesel consumption of a truck per kilometer, expressed in L / km. This indicates the density of diesel fuel, expressed in kg / L. Indicates diesel emission factor; Indicates the rated load capacity of the truck; Indicates the weight of the package; This represents the proportion of the package weight to the vehicle's load capacity, i.e., the two-factor comprehensive allocation coefficient in the DC-DS stage.

[0098] By dividing the process into the three stages described above, this invention can not only calculate overall transportation emissions, but also identify the differences in emission contributions between long-haul trunk transportation, intra-city transshipment, and regional distribution, thereby providing a data foundation for logistics route optimization and transportation tool substitution.

[0099] Carbon emissions from last-mile delivery Last-mile delivery refers to the final mile of delivery from the destination courier station to the consumer's address or pickup point. This stage typically uses electric tricycles for delivery, and the emission calculation formula is as follows: ; in, Indicates package The average short-distance delivery distance from the courier station to the consumer or merchant; This indicates the power consumption of an electric tricycle per kilometer, expressed in kWh / km. The emission factor representing electricity; This indicates the rated cargo capacity of the electric tricycle; Indicates the weight of the package; It represents the proportion of the package weight to the load capacity of the delivery vehicle, and is used for allocation, that is, the two-factor comprehensive allocation coefficient in the last-mile delivery stage.

[0100] For scenarios where the actual delivery distance cannot be obtained, a delivery distance estimation method based on administrative divisions can be used. Specifically, this involves estimating the distance by calling the distance between the regional center point in the geographic system, based on the administrative region (such as street or community) where the courier station and the delivery address are located. The advantage of this alternative is that it eliminates the need to analyze precise latitude and longitude or delivery trajectory for each order; it utilizes public geographic information data to complete the distance estimation, protecting user location privacy while reducing computational complexity.

[0101] The following simulation of the carbon emission accounting process for a standard B2C / C2C online shopping package further illustrates the present invention.

[0102] Item being accounted for: One smart thermos cup shipped from Hangzhou, Zhejiang Province to Urumqi, Xinjiang Uygur Autonomous Region.

[0103] Logistics route: Hangzhou originating express station → Hangzhou provincial transfer center → Urumqi provincial transfer center → Urumqi destination express station → consumer.

[0104] Specific data: Packaging material is corrugated cardboard box; weight of the cardboard box. =0.1kg; Package weight =0.5kg, package volume =0.025m 3 Storage dwell time =12h; Distance from Hangzhou departure point express station to provincial transfer center =35km, intercity transportation will be by truck throughout, the road distance from Hangzhou to Urumqi =3200km, distance from Urumqi Provincial Transfer Center to the destination express station =30km; Electric tricycle is selected as the last-mile delivery method; last-mile delivery distance... =0.8km.

[0105] The first step is to determine the corrugated cardboard box emission factor based on the weight of the packaging materials. =1.14 kg CO2eq / kg), substitute into the formula to calculate the carbon emissions from the packaging process: kgCO2eq; The second step is to determine the package volume, storage time, and lighting power per unit volume. =0.0025kW / m 3 ) and electricity emission factor ( =0.8843 kg CO2eq / kWh), and the emissions from the storage process are calculated using the formula: kgCO2eq; The third step involves calculating the diesel consumption per kilometer of the truck based on the package weight and the transportation distance at different stages. =0.15L / km), density of diesel fuel ( =0.84kg / L), diesel emission factor ( =3.15kgCO2eq / kg) and the rated load capacity of the truck ( =10t), and the emissions from intercity trunk line transportation are calculated according to the formula as follows:

[0106] kgCO2eq

[0107] kgCO2eq

[0108] kgCO2eq

[0109] kgCO2eq The fourth step involves determining the electric tricycle's power consumption per kilometer based on the final delivery distance and package weight. =0.03kWh / km), rated cargo weight ( =200kg) and electricity emission factor ( =0.8843 kg CO2eq / kWh), and the emissions from the last-mile delivery stage are calculated according to the formula: kgCO2eq The fifth step is to aggregate the total emissions across the entire process, adding up the emissions from packaging, warehousing, intercity trunk transportation, and last-mile delivery to obtain the total emissions for the package. Greenhouse gas emissions throughout the entire process: kgCO2eq Based on the method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion provided in the above embodiments, this application also provides a specific implementation of a device for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion. Please refer to the following embodiments.

[0110] First see Figure 2 The carbon emission accounting device 200 for cross-regional commodity circulation based on multi-source big data fusion provided in this application embodiment may include: Module 201 is used to acquire multi-source heterogeneous big data when the target product is circulated across regions; The decomposition module 202 is used to decompose the cross-regional circulation of the target product into multiple carbon emission circulation links, including packaging, warehousing, intercity trunk transportation and last-mile delivery. Module 203 is used to define the carbon emissions of each stage as endogenous variables and the exogenous factors affecting the carbon emissions of each stage as exogenous variables. Based on the physical process of cross-regional circulation of goods, a causal graph containing directed edges of causal transmission between endogenous and exogenous variables is constructed. The causal graph contains at least one directed edge of causal transmission from the carbon emissions of the warehousing stage to the carbon emissions of the intercity trunk line transportation stage, so as to represent the causal transmission relationship that changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. The determination module 204 is used to dynamically determine the weight weight and volume weight based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle in the intercity trunk transportation and last-mile delivery links. The first calculation module 205 is used to calculate the two-factor comprehensive allocation coefficient for each package based on weight weight and volume weight. The second calculation module 206 is used to establish a causal transmission path from the packaging variable through the package weight and package volume to the two-factor comprehensive allocation coefficient and then to the carbon emissions in the transportation stage, using the two-factor comprehensive allocation coefficient as an intermediate node in the causal graph. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated. The fused causal graph includes the causal transmission path.

[0111] Figure 3 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0112] An electronic device may include a processor 301 and a memory 302 storing computer program instructions.

[0113] Specifically, the processor 301 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0114] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. In one instance, memory 302 may include removable or non-removable (or fixed) media, or memory 302 may be a non-volatile solid-state memory. Memory 302 may be internal or external to an electronic device.

[0115] In one instance, memory 302 may be read-only memory (ROM). In one instance, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0116] Memory 302 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion according to the first aspect of this disclosure.

[0117] The processor 301 reads and executes computer program instructions stored in the memory 302 to achieve... Figure 1 The embodiment shown is a method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion.

[0118] In one example, the electronic device may also include a communication interface 303 and a bus 304. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 304 and complete communication with each other.

[0119] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0120] Bus 304 includes hardware, software, or both, that couples components of an electronic device together. For example, and not as a limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 304 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0121] This electronic device can execute the carbon emission accounting method for cross-regional commodity circulation based on multi-source big data fusion as described in the embodiments of this application, thereby achieving a combination of... Figures 1-2 The invention describes a method and apparatus for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion.

[0122] Furthermore, in conjunction with the multi-source big data fusion-based carbon emission accounting method for cross-regional commodity circulation described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the multi-source big data fusion-based carbon emission accounting methods for cross-regional commodity circulation described in the above embodiments.

[0123] In an optional embodiment, in conjunction with the multi-source big data fusion-based carbon emission accounting method for cross-regional commodity circulation described in the above embodiments, this application embodiment can provide a computer program product to implement it. The instructions in the computer program product are executed by the processor of an electronic device, enabling the electronic device to implement any of the multi-source big data fusion-based carbon emission accounting methods for cross-regional commodity circulation described in the above embodiments.

[0124] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0125] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0126] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0127] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0128] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for calculating carbon emissions from cross-regional commodity circulation based on multi-source big data fusion, characterized in that, include: Acquire multi-source heterogeneous big data during the cross-regional circulation of target products; The cross-regional circulation of the target product is broken down into multiple carbon emission circulation links, including packaging, warehousing, intercity trunk transportation, and last-mile delivery. Carbon emissions at each stage are defined as endogenous variables, and exogenous factors affecting carbon emissions at each stage are defined as exogenous variables. A causal graph containing directed edges of causal transmission between endogenous and exogenous variables is constructed based on the physical process of cross-regional commodity circulation. The causal graph contains at least one directed edge of causal transmission from carbon emissions at the warehousing stage to carbon emissions at the intercity trunk transportation stage, in order to characterize the causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. In the intercity trunk line transportation and last-mile delivery links, weight weight and volume weight are dynamically determined based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle. A two-factor comprehensive allocation coefficient is calculated for each package based on the weight weight and volume weight. Using the dual-factor comprehensive allocation coefficient as an intermediate node in the causal graph, a causal transmission path is established from the packaging stage variable through package weight and package volume to the dual-factor comprehensive allocation coefficient and then to the carbon emissions in the transportation stage. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated; wherein, the fused causal graph includes the causal transmission path.

2. The method according to claim 1, characterized in that, The acquisition of multi-source heterogeneous big data during the cross-regional circulation of target products includes: Obtain a product attribute dataset from an e-commerce platform, wherein the product attribute dataset includes at least product category, product sales volume, package volume, package weight, and shipping address; Obtain a packaging material dataset, which includes at least the type of packaging material used in the package, the weight of the packaging material, and the emission factor of the corresponding packaging material. Obtain a warehouse dataset, which includes at least the dwell time of goods in warehouses, distribution centers or delivery centers, the volume occupied by packages, the lighting power per unit volume and the power emission factor; Obtain a transportation vehicle parameter dataset, which includes at least the energy consumption per unit distance, energy density, rated cargo weight, rated cargo volume, and corresponding energy emission factor for each transportation vehicle. Obtain a transportation distance dataset, which includes at least the distance from the originating express station to the originating provincial transfer center, the distance from the originating provincial transfer center to the destination provincial transfer center, the distance from the destination provincial transfer center to the destination express station, and the last-mile delivery distance; Obtain a regional parameter dataset, which includes at least the regional electricity emission factors between different provinces or cities; Data from different data sources is standardized and converted according to a unified unit of measurement to obtain multi-source heterogeneous big data, in which volume is unified to cubic meters, weight to kilograms, distance to kilometers, energy consumption to liters per kilometer or kilowatt-hours per kilometer, and emissions to kilograms of carbon dioxide equivalent.

3. The method according to claim 1, characterized in that, The construction of a causal graph based on the physical process of cross-regional commodity circulation, including directed edges for causal transmission between endogenous and exogenous variables, includes: Based on the physical process of cross-regional commodity circulation, the following causal transmission directed edges are constructed in the causal graph: The directed edge from the variable of packaging specification selection to the carbon emissions of the packaging process represents the direct causal impact of packaging material type and usage on carbon emissions of the packaging process. By selecting the variable of packaging specifications and pointing to the directed edges of package volume and weight to the carbon emissions of warehousing and intercity trunk transportation, we can characterize the indirect causal impact of packaging specifications on the space occupancy of warehousing and the allocation coefficient of transportation by changing package volume and weight. The directed edge from the warehousing strategy variable to the warehousing dwell time variable and then to the carbon emissions in the warehousing process represents the direct causal impact of warehousing strategy on carbon emissions in the warehousing process by influencing the parcel dwell time. The directed edge from carbon emissions in the warehousing stage to carbon emissions in the intercity trunk transportation stage represents the cross-stage causal transmission relationship in which changes in warehousing dwell time affect the transportation batch organization method and loading rate, thereby leading to changes in transportation emissions. Directed edges pointing from weather and road condition variables to carbon emissions in intercity trunk transportation represent the direct causal impact of external environmental factors on transportation efficiency and energy consumption. Directed edges pointing to the total carbon emissions from packaging, warehousing, intercity trunk transportation, and last-mile delivery represent the causal convergence relationship where emissions from each stage are aggregated into the total emissions.

4. The method according to claim 3, characterized in that, After constructing a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation, the method further includes: For each endogenous variable, a structural equation is established with its parent variable as the independent variable. The structural equation is used to quantify the magnitude of the causal effect of each parent variable on the endogenous variable. The structural equations include at least a structural equation for carbon emissions in the intercity trunk transportation segment. The independent variables of the structural equations include at least weather condition variables, road condition variables, carbon emissions in the warehousing segment, and packaging specification variables, in order to quantify the causal influence of the warehousing segment on the transportation segment through the causal transmission directed edge, as well as the causal influence of external environmental factors on the transportation segment.

5. The method according to claim 1, characterized in that, In the intercity trunk line transportation and last-mile delivery stages, weight weights and volume weights are dynamically determined based on the real-time weight and volume utilization rates of the transportation vehicles, including: Obtain the total weight of all packages currently loaded on the transport vehicle, and calculate the ratio of the total weight to the rated cargo capacity of the transport vehicle as the weight utilization rate of the current load; Obtain the total volume of all packages currently loaded on the transport vehicle, and calculate the ratio of the total volume to the rated cargo volume of the transport vehicle as the current volume utilization rate; The weight utilization rate is defined as the proportion of the weight utilization rate in the sum of the weight utilization rate and the volume utilization rate as the weight weight, and the volume utilization rate is defined as the proportion of the volume utilization rate in the sum of the weight utilization rate and the volume utilization rate as the volume weight.

6. The method according to claim 1, characterized in that, The calculation of the two-factor comprehensive allocation coefficient for each package based on the weight weight and volume weight includes: Obtain the total weight and total volume of all packages in the current transportation batch, and calculate the weight of each package as the proportion of the total weight and the volume of each package as the proportion of the total volume. Multiply the weight weight by the weight percentage of the package to obtain the weight distribution of the package; Multiply the volume weight by the volume percentage of the package to obtain the volume allocation component of the package; The weight allocation component and the volume allocation component are added together, and the sum is used as the two-factor comprehensive allocation coefficient for the package in the transportation batch.

7. The method according to claim 1, characterized in that, After constructing a causal graph containing directed edges for causal transmission between endogenous and exogenous variables based on the physical process of cross-domain commodity circulation, the method further includes: The exogenous variable is set to a first value through intervention, while keeping other exogenous variables unchanged. The values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the first expected value of the target variable under the first value. The exogenous variable is set to a second value through intervention, while keeping other exogenous variables unchanged. The values ​​of all endogenous variables are recalculated according to the topological order of the causal graph to obtain the second expected value of the target variable under the second value. The difference between the first expected value under the first value and the second expected value under the second value is calculated as the total causal effect of the exogenous variable on the target variable, wherein the total causal effect includes direct causal effects and indirect causal effects: The direct causal effect is the causal influence of the exogenous variable directly pointing to the target variable through the directed edge in the causal graph. It is obtained by calculating the intervention effect after blocking all other paths except the direct directed edge in the causal graph. The indirect causal effect is the causal influence of the exogenous variable transmitted to the target variable through a directed path via at least one intermediate variable in the causal graph, and is obtained by subtracting the direct causal effect from the total causal effect.

8. The method according to claim 7, characterized in that, After calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: When the value of any exogenous variable changes, trace the complete path of that change propagating node by node along each causal directed edge in the causal graph; Record the increment of causal effect produced as the change propagates along each directed edge of causal transmission; The incremental causal effects are aggregated according to their respective causal transmission paths, and the effect proportion of each causal transmission path is output to identify the dominant causal transmission path and key intermediate variables that affect the total carbon emissions from this exogenous variable.

9. The method according to claim 7, characterized in that, After calculating the difference between the first expected value under the first value and the second expected value under the second value as the total causal effect of the exogenous variable on the target variable, the method further includes: Obtain the actual observation data of the target order, and record the actual values ​​of each exogenous variable and the actual observed values ​​of each endogenous variable; The value of the target exogenous variable is changed from the actual observed value to a preset counterfactual value, while keeping other exogenous variables unchanged; Following the topological order of the causal graph, starting from the exogenous variables, the values ​​of all endogenous variables are recalculated node by node to obtain the carbon emissions of each stage and the total carbon emissions under the counterfactual scenario. By comparing the carbon emissions at each stage under the counterfactual scenario with the actual observed values, the causal effect of the change in this exogenous variable on the total carbon emissions can be quantified.

10. A carbon emission accounting device for cross-regional commodity circulation based on multi-source big data fusion, characterized in that, The device includes: The acquisition module is used to acquire multi-source heterogeneous big data when the target product is distributed across regions; The decomposition module is used to decompose the cross-regional circulation of the target product into multiple carbon emission circulation links, including packaging, warehousing, intercity trunk transportation, and last-mile delivery. The module is used to define the carbon emissions of each stage as endogenous variables and the exogenous factors affecting the carbon emissions of each stage as exogenous variables. Based on the physical process of cross-regional circulation of goods, a causal graph is constructed, which includes directed edges of causal transmission between endogenous and exogenous variables. The causal graph contains at least one directed edge of causal transmission from the carbon emissions of the warehousing stage to the carbon emissions of the intercity trunk transportation stage, so as to characterize the causal transmission relationship that changes in warehousing dwell time affect the transportation batch organization method and thus lead to changes in transportation emissions. The determination module is used to dynamically determine weight weight and volume weight based on the real-time weight utilization rate and volume utilization rate of the transportation vehicle in the intercity trunk transportation and last-mile delivery stages. The first calculation module is used to calculate the two-factor comprehensive allocation coefficient for each package based on the weight weight and volume weight. The second calculation module is used to establish a causal transmission path from the packaging stage variable through the package weight and package volume to the two-factor comprehensive allocation coefficient and then to the carbon emissions of the transportation stage, using the two-factor comprehensive allocation coefficient as an intermediate node in the causal graph. Based on the fused causal graph, the carbon emissions of each stage and the total carbon emissions are calculated. The fused causal graph includes the causal transmission path.

Images