Climate benefit-driven project full-cycle carbon footprint management method and system

Abstract

The invention discloses a climate-benefit-driven project full-cycle carbon footprint management method and system, and relates to the technical field of carbon footprint management and carbon emission accounting, and the method comprises the steps: obtaining enterprise project full-cycle multiple processes, separately establishing direct and indirect carbon emission management units, and synthesizing the direct and indirect carbon emission management units, and constructing a dual-drive climate benefit measurement and calculation model containing double sub-models, measuring and calculating to obtain direct and indirect measurement and calculation indexes, calculating a comprehensive measurement and calculation index, and sending the comprehensive measurement and calculation index to a climate investment and financing platform for storage. The technical problems that in carbon emission transaction, a traditional carbon footprint management method cannot finely distinguish direct and indirect carbon emission units in the whole period of a project, indirect emission such as a supply chain is difficult to accurately measure and calculate, and carbon footprint data is not comprehensive and accurate are solved. The technical effects of fine quantification of the full-cycle carbon footprint of the project and accurate measurement and calculation of climate benefits are achieved, and comprehensive and accurate data support is provided for project evaluation and carbon asset pricing in carbon emission transaction.

Description

Technical Field

[0001] This invention relates to the field of carbon footprint management and carbon emission accounting technology, and in particular to a climate benefit-driven method and system for full-cycle carbon footprint management of projects. Background Technology

[0002] In the carbon emissions trading process, accurate carbon footprint management is crucial for project evaluation and trading decisions. Current technologies primarily employ traditional single-modal data acquisition and extensive calculation methods to detect the carbon footprint of a project throughout its entire lifecycle. While these methods have proven effective in closed or stable environments, their limitations have become apparent when applied to full-lifecycle project management as climate investment markets demand higher precision in carbon data.

[0003] Because the entire project lifecycle involves multiple processes and stages, traditional methods cannot accurately distinguish between direct and indirect carbon emission units, and lack effective calculations of indirect emissions from the supply chain, resulting in incomplete and inaccurate carbon footprint data. This makes it difficult to meet the needs of accurate assessment and effective management of project climate benefits in carbon emission trading, and also fails to provide a reliable basis for carbon asset pricing and trading. Summary of the Invention

[0004] This application provides a climate benefit-driven project lifecycle carbon footprint management method and system to address the technical problem that traditional carbon footprint management methods in carbon emission trading cannot accurately distinguish between direct and indirect carbon emission units throughout the project lifecycle, and are difficult to accurately measure indirect emissions such as those from the supply chain, resulting in incomplete and inaccurate carbon footprint data.

[0005] The first aspect of this application provides a climate benefit-driven method for managing the full lifecycle carbon footprint of a project. The method includes: acquiring multiple project processes throughout the entire project lifecycle; separating the multiple project processes to establish direct carbon emission management units and indirect carbon emission management units for each project process; synthesizing multiple direct carbon emission management units and multiple indirect carbon emission management units from the multiple project processes; constructing a dual-drive climate benefit calculation model, which includes a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model; calculating the multiple direct emission management units using the direct-drive climate benefit calculation sub-model, and calculating the multiple indirect emission management units using the indirect-drive climate benefit calculation sub-model to obtain direct and indirect calculation indicators; calculating a comprehensive calculation indicator of the direct and indirect calculation indicators, and sending the comprehensive calculation indicator to a climate investment and financing platform for storage.

[0006] A second aspect of this application provides a climate benefit-driven project lifecycle carbon footprint management system, comprising: a project process acquisition module for acquiring multiple project processes throughout the entire project lifecycle of an enterprise; a project process separation module for separating the multiple project processes, establishing direct carbon emission management units and indirect carbon emission management units for each project process, and synthesizing multiple direct carbon emission management units and multiple indirect carbon emission management units for the multiple project processes; a climate model construction module for constructing a dual-drive climate benefit calculation model, the dual-drive climate benefit calculation model including a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model; a calculation index acquisition module for calculating multiple direct emission management units using the direct-drive climate benefit calculation sub-model, and calculating multiple indirect emission management units using the indirect-drive climate benefit calculation sub-model, to obtain direct calculation indicators and indirect calculation indicators; and a calculation index storage module for calculating a comprehensive calculation index of the direct calculation indicators and the indirect calculation indicators, and sending the comprehensive calculation index to a climate investment and financing platform for storage.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] This application acquires carbon footprint activity data from multiple processes throughout the entire project lifecycle, separates direct and indirect carbon emission management units, constructs a dual-drive climate benefit calculation model, and calculates comprehensive measurement indicators after multimodal data classification, factor matching, and indicator system training. It then dynamically optimizes the model based on third-party feedback, thereby accurately quantifying the direct and indirect carbon emissions and climate benefits throughout the entire project lifecycle. This makes the carbon footprint management and benefit assessment results of climate investment and financing projects more scientific and reliable, achieving refined quantification of the project's carbon footprint and accurate calculation of climate benefits throughout the entire project lifecycle. This provides comprehensive and accurate data support for project evaluation and carbon asset pricing in carbon emission trading. Attached Figure Description

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

[0010] Figure 1 This is a flowchart illustrating a climate benefit-driven project lifecycle carbon footprint management method provided in this application embodiment.

[0011] Figure 2 This is a schematic diagram of a climate benefit-driven project lifecycle carbon footprint management system provided in an embodiment of this application.

[0012] Figure labeling: Project process acquisition module 1, Project process separation module 2, Climate model construction module 3, Calculation index acquisition module 4, Calculation index storage module 5. Detailed Implementation

[0013] This application provides a climate benefit-driven project lifecycle carbon footprint management method and system to address the technical problem that traditional carbon footprint management methods in carbon emission trading cannot accurately distinguish between direct and indirect carbon emission units throughout the project lifecycle, and are difficult to accurately measure indirect emissions such as those from the supply chain, resulting in incomplete and inaccurate carbon footprint data.

[0014] 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 a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0015] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0016] Example 1, as Figure 1 As shown, a climate benefit-driven approach to project lifecycle carbon footprint management includes:

[0017] Step A100: Obtain multiple project processes throughout the entire lifecycle of the enterprise project.

[0018] Specifically, in the context of carbon emission management, acquiring multiple project processes throughout the entire lifecycle of a company's project refers to data collection and process streamlining at each stage of the project's complete lifecycle, from planning to operation. Specifically, this needs to cover raw material supply chain information during the project planning phase (such as upstream enterprise capacity and transportation routes), construction energy consumption data during the construction phase (such as electricity consumption of construction machinery and building material consumption), real-time energy consumption monitoring during the production and operation phase (such as hourly electricity consumption of production equipment and steam usage), and material flow records throughout the product's lifecycle (such as raw material input and finished product output). This data comes from multiple sources, including the company's production system, supply chain management platform, and regulatory databases. For example, it connects data from departments such as taxation and industrial added value through big data center interfaces, while simultaneously utilizing IoT devices to capture real-time operating parameters of production equipment, such as the raw coal consumption and power generation per shift on a particular production line.

[0019] In terms of data format, the acquired project process information includes multimodal features, such as text-based supplier contract documents, numerical energy consumption statistics, and time-series quarterly emission monitoring records. Taking an industrial technological upgrading project as an example, it is necessary to collect annual energy consumption reports for the three years prior to the upgrade (e.g., annual standard coal consumption of 1000 tons), energy-saving monitoring data for each stage after the upgrade (e.g., standard coal consumption reduced to 800 tons in the first year after the upgrade), and loan fund flow data throughout the project lifecycle (e.g., a 2 million yuan loan corresponding to the production line upgrade stage). Through the above methods, a digital mapping of each process node throughout the project lifecycle is achieved, ensuring that the data covers both direct production activities (e.g., workshop energy consumption) and indirectly related stages (e.g., carbon emissions from raw material transportation).

[0020] By collecting multimodal data from various processes throughout the entire project lifecycle through multiple sources, and combining big data interfaces with IoT real-time monitoring technology, the project achieved comprehensive acquisition of carbon footprint activity data throughout its entire lifecycle, providing a complete data foundation for the subsequent separation of direct and indirect carbon emission management units.

[0021] Step A200: Separate the multiple project processes, establish a direct carbon emission management unit and an indirect carbon emission management unit for each project process, and synthesize multiple direct carbon emission management units and multiple indirect carbon emission management units for the multiple project processes.

[0022] In this embodiment, the direct carbon emission management unit is a carrier for collecting carbon emission data generated by the enterprise's autonomous and controllable activities in the project process, focusing on directly related links such as production equipment operation and process combustion. The indirect carbon emission management unit is a carrier for aggregating carbon emission data indirectly caused by the enterprise's activities in the project process, covering related links such as raw material procurement, supply chain transportation, and upstream production.

[0023] Optionally, multiple project processes can be separated first. Taking chemical project A (raw material procurement → production → emission monitoring), photovoltaic project B (silicon procurement → module production → installation), and logistics project C (goods transportation → distribution) as examples, the independent boundaries of each project process can be clarified by analyzing the activity descriptions of the entire project lifecycle, such as equipment operation in A, module production in B, and transportation and distribution in C; location information, such as within the factory area of ​​A, within the factory area of ​​B, and outside the factory area of ​​C; and industry codes, such as manufacturing for A, energy for B, and services for C. This avoids confusion and overlap between cross-project processes.

[0024] Based on the separated project processes, direct and indirect carbon emission management units are established for each project process. Specific steps are detailed in A210-A220. For Project A, the direct carbon emission management unit records numerical characteristics such as energy consumption and emissions from boilers in the production workshop, as well as categorized characteristics such as emission source type and equipment type. The indirect carbon emission management unit records spatiotemporal and textual characteristics such as location information and supply chain hierarchy in the raw material procurement process. For Project B, the direct unit focuses on the power consumption and activity intensity of motors in the module production line, while the indirect unit covers upstream energy consumption in silicon material production. For Project C, the direct unit collects fuel consumption and emissions from transportation vehicles, while the indirect unit records implicit emissions from vehicle manufacturing.

[0025] Subsequently, multiple direct carbon emission management units and multiple indirect carbon emission management units are synthesized from the aforementioned project processes. In the direct emission dimension, boiler emissions from Project A, motor emissions from Project B, and vehicle emissions from Project C are aggregated. Through numerical feature overlay, such as 5000 + 2000 + 1500 = 8500 tons of CO2 equivalent, and combined with equipment type characteristics such as boiler / motor / vehicle categories, a cross-project direct emission pool is constructed, clearly presenting the proportion of direct emissions for each project and the equipment distribution. In the indirect emission dimension, supplier transportation data from Project A, upstream silicon material data from Project B, and vehicle manufacturing data from Project C are integrated. Based on location information, supply chain hierarchy, and other spatiotemporal characteristics, as well as textual features such as raw material procurement and upstream processing, an indirect emission pool is formed, accurately locating the supply chain links and responsible entities for indirect emissions.

[0026] By separating multiple project processes and constructing direct / indirect emission units, and aggregating similar data, a structured view of direct and indirect carbon emissions across projects was achieved, providing macro-level data support for the overall planning of group-level carbon assets and the calculation of climate benefits.

[0027] Step A300: Construct a dual-drive climate benefit calculation model, which includes a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model.

[0028] In one embodiment of this application, constructing a sub-model for measuring direct-driven climate benefits requires identifying direct carbon footprint activity data samples of direct carbon emission management units, matching direct carbon emission factors to calculate total direct carbon emissions, identifying direct-energy-saving carbon footprints and corresponding emission reductions, and training the sub-model based on a climate benefit index system. The specific steps are detailed in A310-A340.

[0029] To construct a sub-model for measuring indirect climate benefits, it is necessary to identify data samples of indirect carbon footprint activities of indirect carbon emission management units, obtain indirect carbon emission factors to calculate total indirect carbon emissions, identify indirect energy-saving carbon footprints and corresponding emission reductions, and train the sub-model based on the same climate benefit indicator system. The specific steps are explained in detail in A350-A380.

[0030] Step A400: Calculate the direct-driven climate benefit calculation sub-model for multiple direct emission management units, and calculate the indirect-driven climate benefit calculation sub-model for multiple indirect emission management units to obtain direct and indirect calculation indicators.

[0031] Specifically, when using the direct-driven climate benefit calculation sub-model to calculate the direct emissions management unit, the first input is a sample of direct carbon footprint activity data, such as a power plant boiler consuming 5,000 tons of coal and a motor consuming 2 million kilowatt-hours of electricity annually, along with corresponding direct carbon emission factor samples, such as a coal combustion factor of 1.8 kg CO2 / kg coal and an electricity factor of 0.8 kg CO2 / kWh. The calculated total direct carbon emissions sample is 5,000 tons × 1,000 kg / ton × 1.8 kg CO2 / kg coal + 200 × 10,000 kWh × 0.8 kg CO2 / kWh = 10,600,000 kg CO2 = 10,600 tons of CO2.

[0032] Simultaneously, the direct-energy carbon footprint data is identified, such as the annual coal consumption decreasing to 4500 tons after boiler retrofitting. The direct-carbon emission reduction is calculated as (5000-4500) tons × 1000 kg / ton × 1.8 kg CO2 / kg coal = 900 tons of CO2. Next, based on the climate benefit index system, the direct emission reduction contribution value of 900 tons, the energy efficiency improvement coefficient (5000-4500) / 5000 = 0.1, the resource recycling impact weight (if the waste recycling rate increases by 10%, the weight becomes 0.2), and the environmental synergy improvement coefficient (a reduction of 20 tons of sulfur dioxide is converted to a CO2 equivalent of 200 tons, coefficient 200 / 900≈0.22). These parameters are input into the model, and the directly measured index is output (such as a climate benefit score of 85 points).

[0033] For the indirect climate benefit calculation model, taking a car company's chip transportation as an example, indirect carbon footprint data is first obtained. The transportation distance is 500 kilometers, and the fuel consumption is 15 liters / 100 kilometers. Through the indirect conversion module, the direct carbon emission factor (diesel combustion 2.3 kg CO2 / L) is converted into an indirect factor of 2.3 × 1.3 = 2.99 kg CO2 / L, with an empty load rate of 20% and a packaging weight impact of 10%. The total indirect carbon emissions are calculated to be 500 / 100 × 15 × 2.99 = 224.25 kg CO2. If electric trucks are subsequently used, the fuel consumption will be reduced to 10 liters / 100 kilometers, and the reduction in indirect carbon emissions will be calculated to be (15-10) × 5 × 2.99 = 74.75 kg CO2. Combining the climate benefit index system, the indirect emission reduction contribution value is calculated as 74.75 kg, the energy efficiency improvement coefficient is (15-10) / 15≈0.33, the supply chain impact weight is 0.4 (Tier 1 supplier), and the environmental synergy coefficient is calculated (1 kg of nitrogen oxide emission reduction is converted to 10 kg of CO2, coefficient 10 / 74.75≈0.13). After inputting these values ​​into the model, the indirect measurement index is output (such as a climate benefit score of 78 points).

[0034] By connecting the direct and indirect emission management units to their respective driving sub-models, and combining a multi-dimensional indicator system and data conversion mechanism, the accurate quantification of the project's carbon footprint throughout its entire lifecycle was achieved, providing a scientific underlying support for the generation of comprehensive calculation indicators for the climate investment and financing platform.

[0035] Step A500: Calculate the comprehensive measurement index of the direct measurement index and the indirect measurement index, and send the comprehensive measurement index to the climate investment and financing platform for storage.

[0036] In this embodiment of the application, the climate investment and financing platform is a systematic platform used to store comprehensive calculation indicators generated from the full life cycle carbon footprint management of projects and to provide climate investment and financing related services to enterprises, institutions and other entities.

[0037] Specifically, when calculating the comprehensive assessment indicators, it is first necessary to obtain both direct and indirect assessment indicators. Taking a power plant as an example, the direct-driven climate benefit assessment sub-model outputs the following direct assessment indicators: direct emission reduction contribution of 900 tons of CO2, energy efficiency improvement coefficient of 0.1, resource recycling impact weight of 0.2, and environmental synergy improvement coefficient of 0.22, which, after standardization, results in a score of 85. The indirect-driven climate benefit assessment sub-model, targeting the chip transportation scenario, outputs the following indirect assessment indicators: indirect emission reduction contribution of 74.75 kg of CO2, energy efficiency improvement coefficient of 0.33, supply chain impact weight of 0.4, and environmental synergy coefficient of 0.13, which, after standardization, results in a score of 78.

[0038] Subsequently, a weighting system was determined based on industry characteristics and expert experience, such as a 60% weighting for direct indicators and a 40% weighting for indirect indicators in the manufacturing sector. A weighted average method was used to calculate the comprehensive evaluation index: Comprehensive Score = Direct Score × 60% + Indirect Score × 40% = 85 × 0.6 + 78 × 0.4 = 82.2 points. Simultaneously, raw data (such as total direct emissions of 10,600 tons and total indirect emissions of 224.25 kg) was integrated with standardized indicators to form a structured data package.

[0039] Finally, the comprehensive calculation indicators are sent to the climate investment and financing platform for storage through the blockchain interface. Smart contracts are used to ensure that the data is tamper-proof, while linking basic enterprise information (such as unified credit code and loan standards) with project database fields (such as project stage and loan balance).

[0040] By constructing a standardized system and weighted integration mechanism for direct and indirect indicators, and combining it with blockchain technology to achieve credible data on-chain, the system has achieved the effect of accurately quantifying the full-cycle climate benefits and providing a scientific basis for investment and financing decisions, thus promoting closed-loop management of data-assessment-decision in climate investment and financing scenarios.

[0041] Furthermore, step A200 in the method provided in this application embodiment includes:

[0042] A210: Establish direct carbon emission management units and indirect carbon emission management units for each project process. The direct carbon emission management units and the indirect carbon emission management units are connected to the carbon emission data acquisition module through a data interface to collect carbon footprint activity data of each carbon emission management unit.

[0043] A220: Wherein, the carbon footprint activity data includes energy consumption monitoring data, material input and output records, production operation data, and environmental emission monitoring data.

[0044] In this embodiment, the data interface is a channel used to connect the direct carbon emission management unit and the indirect carbon emission management unit to the carbon emission data acquisition module to collect carbon footprint activity data for each carbon emission management unit. The carbon emission data acquisition module is used to collect carbon footprint activity data and includes a multimodal fusion classifier that can classify the carbon footprint activity data and output classification labels.

[0045] Specifically, the process begins by breaking down the project into stages based on its entire lifecycle (planning, construction, operation, etc.). Taking a new energy project as an example, the direct carbon emission management unit covers the energy consumption of equipment in the production workshop, such as the electricity consumption of each unit at 2000 kWh per hour and the steam consumption at 50 tons per day; material input and output records, such as the daily raw material input of 100 tons and the finished product output of 80 tons; the indirect management unit corresponds to the production capacity data of raw material suppliers, such as the annual production capacity of upstream enterprises at 100,000 tons; and transportation route energy consumption, such as the fuel consumption of raw material transportation at 15 liters per 100 kilometers. These management units are connected in real time to the carbon emission data acquisition module through data interfaces, using API interfaces to obtain material flow data from the enterprise's ERP system, and collecting real-time energy consumption monitoring data of production equipment through IoT sensors, such as the current and voltage parameters of a certain production line per minute.

[0046] The collection of carbon footprint activity data needs to cover multi-source heterogeneous information. Taking a chemical project as an example, energy consumption monitoring data includes an average annual consumption of 5,000 tons of standard coal before the upgrade, which was reduced to 4,200 tons after the upgrade; material input and output records include a monthly input of 200 tons of raw material styrene and an output of 180 tons of polystyrene; production operation data involves process parameters such as reactor temperature and pressure, for example, the reaction temperature is maintained at 80-90℃ and the pressure is 0.5MPa; environmental emission monitoring data includes the CO2 concentration at the chimney outlet (e.g., a real-time monitoring value of 400ppm) and dust emissions (e.g., 0.5kg per hour). This data is synchronized to the carbon emission data acquisition module through a data interface, for example, obtaining production operation parameters from PLC equipment via the Modbus protocol and obtaining material ledger records through database integration.

[0047] Through the above steps, the structured separation and accurate collection of direct and indirect carbon emission data throughout the project's entire lifecycle were achieved. A unit-based data collection mechanism was established, and combined with real-time data synchronization technology from multiple sources, the carbon footprint activity data was accurately classified according to direct / indirect dimensions, providing a structured data foundation for subsequent dual-drive climate benefit calculation models.

[0048] Furthermore, step A210 in the method provided in this application embodiment includes:

[0049] A211: The carbon emission data acquisition module includes a multimodal fusion classifier, which is used to classify the carbon footprint activity data after receiving it and output at least one classification label, including direct emission labels and indirect emission labels.

[0050] A212: Based on at least one output classification label, the corresponding data is routed and written into the direct carbon emission management unit or the indirect carbon emission management unit.

[0051] In this embodiment, the multimodal fusion classifier is the core component of the carbon emission data acquisition module, used to classify and process carbon footprint activity data.

[0052] Optionally, the carbon emission data acquisition module is equipped with a multimodal fusion classifier. Its workflow is as follows: First, it receives carbon footprint activity data from multiple sources, such as energy consumption monitoring data of photovoltaic panel production in a new energy project (numerical, such as 2000 kWh per hour), raw material silicon wafer transportation records (text, such as a transportation route of 500 km), and production line operating parameters (time series, such as equipment temperature fluctuations from 8:00 to 18:00 daily). The input layer of the classifier parses the route information in the transportation records through a text feature extractor, processes the power consumption data through a numerical feature extractor, and captures the periodic patterns of equipment temperature through a time series feature extractor. Finally, a comprehensive feature vector is generated through a feature fusion layer.

[0053] Taking a chemical project as an example, when the classifier receives reactor energy consumption data (related to direct emissions) and catalyst supplier capacity data (related to indirect emissions), the classification decision layer judges the feature vectors based on the trained model: the real-time power consumption of the reactor (e.g., 3000 kWh / hour) is assigned a "direct emission label," while the implicit carbon emission data of the catalyst production process (e.g., 5000 kWh of electricity consumed per ton of catalyst production) is labeled as an "indirect emission label." After the classifier outputs the labels, the data routing module writes the reactor energy consumption data into the direct carbon emission management unit and the catalyst capacity data into the indirect carbon emission management unit according to the labels, thus achieving structured separation of data.

[0054] By using a multimodal fusion classifier to intelligently classify and label carbon footprint activity data, the system achieves the effect of accurately separating direct and indirect carbon emission data and writing them into the corresponding management units, providing structured and highly accurate data support for subsequent dual-drive climate benefit measurement models.

[0055] Furthermore, step A211 in the method provided in this application embodiment includes:

[0056] A211-1: The multimodal fusion classifier includes an input layer, a feature fusion layer, a classification decision layer, and a label output layer. The input layer includes a text feature extractor, a numerical feature extractor, a category feature encoder, and a time series feature extractor.

[0057] A211-2: Collect historical carbon footprint activity data samples, extract text features, numerical features, category features and spatiotemporal features using the input layer, input them into the feature fusion layer to obtain fused feature samples, and perform multi-stage training on the classification decision layer based on the fused feature samples until a multimodal fusion classifier with a preset accuracy is obtained.

[0058] A211-3: Deploy the trained multimodal fusion classifier to the carbon emission data acquisition module.

[0059] Specifically, the process begins by collecting historical carbon footprint samples covering all feature dimensions: Textual feature dimension: recording activity descriptions (equipment operation / power consumption / raw material procurement), location information (within / outside / supplier location), and responsible entity (company-owned / outsourced / supplier); Numerical feature dimension: collecting emissions (CO2 equivalent tons), energy consumption (standard coal tons), and activity intensity (emissions per unit product), such as a boiler emitting 5000 tons of CO2 annually and consuming 200 kWh of electricity per unit product; Category feature dimension: labeling emission source type (combustion / process / escape / indirect), industry code (manufacturing / energy / services), and equipment type (boiler / motor / vehicle); Spatiotemporal feature dimension: recording timestamps (data collection time), geographical location (GPS coordinates / administrative division), and supply chain level (first-tier / second-tier / third-tier suppliers). This constructs a sample set containing multiple multimodal data points. Due to space limitations, the above examples are not expanded further; specific details should be determined by those skilled in the art based on the actual project.

[0060] Next, the input layer initiates feature extraction: the text feature extractor parses keywords such as "raw material procurement" to generate semantic vectors, such as "transportation from supplier" mapping to indirect emission vectors; the numerical feature extractor normalizes "5000 tons of CO2" to the 0-1 range; the categorical feature encoder performs One-Hot encoding on "emission source = indirect"; the time series feature extractor captures the periodic patterns of timestamps, such as peak season energy consumption fluctuations, and combines GPS coordinates to spatial vectors to extract spatiotemporal features. The four dimensions of features are then spliced ​​together by the fusion layer to form a multi-dimensional fused sample.

[0061] Based on fused samples, the classification decision layer undergoes multi-stage training: the first stage uses 500 directly labeled samples and 500 indirectly labeled samples, optimizing parameters through cross-entropy loss; the second stage introduces hard example mining, strengthening training for easily confused samples such as "outsourced services" and "energy consumption," until the validation set accuracy reaches 95%. After training, the classifier is deployed to the carbon emission data acquisition module. When receiving data such as "activity description = raw material procurement, location = supplier, emission = 100 tons of CO2, supply chain = level one," it can quickly extract four key features, identify indirect emission labels, and achieve accurate classification.

[0062] By constructing a multimodal sample set covering four major feature dimensions, and through input layer feature extraction and fusion training, a high-accuracy multimodal fusion classifier is trained and deployed to the acquisition module, providing accurate model support for the direct and indirect classification of carbon footprint data.

[0063] Furthermore, step A300 in the method provided in this application embodiment includes:

[0064] A310: Identify a sample of direct carbon footprint activity data for the direct carbon emission management unit.

[0065] A320: Match the direct carbon emission factor sample of each type of carbon footprint activity data sample in the direct carbon footprint activity data sample, and calculate the direct carbon footprint activity data based on the direct carbon emission factor sample to obtain the direct-total carbon emission sample.

[0066] A330: Identify the direct-energy-saving carbon footprint activity data sample of the direct carbon footprint activity data sample, and the direct-carbon emission reduction sample corresponding to the direct-energy-saving carbon footprint activity data sample.

[0067] A340: Define a climate benefit index system, and train the climate benefit on the direct carbon emission reduction sample and the direct carbon emission total sample under the climate benefit index system to obtain a direct-driven climate benefit measurement sub-model.

[0068] In this embodiment, direct carbon footprint activity data refers to carbon emission-related data that is directly generated by the enterprise's own production and operation activities throughout the entire project lifecycle and can be directly monitored and accounted for, and is mainly stored in the direct carbon emission management unit. The climate benefit index system is a multi-dimensional framework for quantitatively assessing the climate impact of a project, including a four-dimensional index system of carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits.

[0069] Specifically, firstly, data samples of direct carbon footprint activities from direct carbon emission management units are identified. From these units, multimodal data encompassing text, numerical, categorical, and spatiotemporal features are extracted. For example, at the textual feature level, activity descriptions are filtered to include boiler operation / motor power consumption / process escape, and location information to include records within the plant area. At the numerical feature level, emissions are collected in tons of CO2 equivalent, and energy consumption in tons of standard coal, such as 5000 tons of coal consumed annually by the boiler and 10000 kWh of electricity consumed monthly by the motor. At the categorical feature level, emission source type is labeled as combustion / power consumption / escape, and equipment type as boiler / motor / reactor. At the spatiotemporal feature level, timestamps are extracted to include monthly data from the past three years, and supply chain level as none (direct activity). Through feature filtering, a direct carbon footprint dataset containing 1000 samples is constructed, with each sample associated with equipment type, emission type (direct), and corresponding parameters.

[0070] Secondly, the direct carbon emission factor samples are matched and the total direct carbon emissions are calculated. Traditional technologies rely on industry average factors, such as a uniform boiler combustion factor of 1.8 tons of CO2 / ton of standard coal. This step precisely matches the equipment and activity types: For example, the annual coal consumption of a boiler is 5,000 tons, matched with a direct carbon emission factor of 1.8 tons of CO2 / ton of standard coal, calculating emissions of 5,000 × 1.8 = 9,000 tons; the monthly electricity consumption of a motor is 10,000 kWh, matched with a direct carbon emission factor of 0.8 kg CO2 / kWh, resulting in monthly emissions of 8 tons and an annual cumulative emissions of 96 tons; the annual leakage of a reactor due to process escape is 100 kg, matched with a factor of 1 kg CO2 / kg (leakage is directly counted as CO2), resulting in annual emissions of 100 kg, or 0.1 tons of CO2. By summarizing the emissions of each type, a sample of the total direct carbon emissions is obtained. For example, this project has an annual direct CO2 emission of 9,096.1 tons of CO2, achieving a structured calculation from multimodal data to the total amount.

[0071] Next, we identified direct-energy-saving carbon footprint activity data and direct-carbon emission reduction samples. From the direct-carbon footprint activity data samples, we screened comparative data after energy-saving retrofits: after boiler retrofitting, annual coal consumption decreased to 4500 tons, reducing emissions by 500 × 1.8 = 900 tons; replacing the motor with a high-efficiency model reduced monthly electricity consumption to 8000 kWh, reducing monthly emissions by 1.6 tons, and annually by 19.2 tons; and the reactor sealing was upgraded, reducing annual leakage to 50 kg, reducing emissions by 0.05 tons. We labeled these retrofitted activity descriptions = energy-saving boiler operation / high-efficiency motor electricity consumption data as direct-energy-saving carbon footprint activity data samples, calculated emission reductions (before retrofitting - after retrofitting), and formed direct-carbon emission reduction samples. For example, this project reduced emissions by 919.25 tons annually, capturing the dynamic contribution of energy-saving measures.

[0072] Finally, under the climate benefit index system (including carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits), the direct emission reduction contribution value, energy efficiency improvement coefficient, influence weight with the total carbon emission sample, and environmental synergy improvement coefficient are calculated respectively. Based on these four types of parameters, a sub-model for directly driving climate benefit measurement is constructed. The specific steps are explained in detail in A341-A346.

[0073] By precisely breaking down the direct emission structure, matching dynamic factors, and constructing multi-dimensional indicator training, the effect of quantifying the contribution of direct carbon emissions to climate benefits has been achieved.

[0074] Furthermore, step A340 in the method provided in this application embodiment includes:

[0075] A341: The climate benefit indicator system includes carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits.

[0076] A342: Calculate the direct emission reduction contribution value of the carbon emission reduction sample based on the carbon emission reduction benefits.

[0077] A343: Calculate the energy efficiency improvement coefficient corresponding to the carbon emission reduction sample based on the energy efficiency improvement benefits.

[0078] A344: Calculate the influence weights of the carbon emission reduction sample and the total carbon emission sample based on the resource recycling benefits.

[0079] A345: Calculate the environmental synergistic improvement coefficient of the carbon emission reduction sample based on the environmental synergistic benefits.

[0080] A346: Using the direct emission reduction contribution value, the energy efficiency improvement coefficient, the impact weight, and the synergistic improvement coefficient as training parameters, a sub-model for measuring direct climate benefits is constructed.

[0081] Specifically, firstly, the composition of the climate benefit indicator system should be clearly defined, including carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits. Taking a chemical project as an example, the carbon emission reduction benefit corresponds to an annual CO2 emission reduction of 919.25 tons after the renovation, which is combined with the emission reduction from the aforementioned steps of upgrading the A330 boiler, motor, etc.; the energy efficiency improvement benefit is related to the reduction in energy consumption per unit of product. Before the renovation, the unit product power consumption was 200 kWh / unit, and after the renovation, it dropped to 160 kWh / unit, a 20% reduction in energy consumption; the resource recycling benefit is reflected in the amount of waste recovered, with the annual amount of waste recovered from the reactor increasing from 50 tons to 80 tons, an increase in recovery rate of 37.5%; the environmental synergy benefit covers the reduction of other pollutants, such as the annual reduction of sulfur dioxide from 20 tons to 5 tons and nitrogen oxides from 15 tons to 3 tons.

[0082] Secondly, training parameters are calculated based on various benefit indicators. According to the carbon emission reduction benefit, the direct emission reduction contribution is 919.25 tons of CO2, quantifying the project's direct improvement on the climate. In the energy efficiency improvement benefit, the energy efficiency improvement coefficient = (energy consumption before modification - energy consumption after modification) / energy consumption before modification = (200 - 160) / 200 = 0.2, reflecting the improvement in energy utilization efficiency. In the resource recycling benefit, the impact weight = (recycling amount / total consumption) × correction coefficient. The project's total waste consumption is 200 tons, and the recycling amount is 80 tons. The weight = (80 / 200) ×1.2 (correction coefficient) = 0.48, reflecting the contribution of resource recycling to climate benefits; in the environmental synergy benefits, the environmental synergy improvement coefficient = (reduction of traditional pollutants / reduction of CO2 emissions) × comprehensive factor. The project's total reduction of sulfur dioxide and nitrogen oxides is (20+15)-(5+3) = 27 tons, which can be converted into CO2 equivalent of about 270 tons at a 10:1 ratio. The improvement coefficient = 270 / 919.25 × 0.8 = 0.235, quantifying the synergistic benefits of other pollutant emission reductions on climate.

[0083] Finally, the direct emission reduction contribution (919.25 tons), energy efficiency improvement coefficient (0.2), impact weight (0.48), and synergistic improvement coefficient (0.235) were used as training parameters and input into a machine learning model (such as a gradient boosting tree) for iterative training. The model uses the climate benefit assessment results of historical projects as labels. For example, if a project has a comprehensive score of 85, the parameter weights are optimized through 1000 iterations to ensure that the prediction errors of both the training and validation sets are less than 5%. Finally, a sub-model for directly driving climate benefit measurement is constructed. The input of this model is multi-dimensional parameters such as carbon emission reduction and energy consumption data, and the output is a comprehensive climate benefit score.

[0084] By constructing an indicator system encompassing four major benefits and calculating multidimensional training parameters, the project's comprehensive climate impact was fully quantified, providing a scientific quantitative basis for green credit assessment and carbon incentive applications.

[0085] Furthermore, step A300 in the method provided in this application embodiment includes:

[0086] A350: Identify a sample of indirect carbon footprint activity data for the indirect carbon emission management unit.

[0087] A360: Obtain the indirect carbon emission factor sample for each type of carbon footprint activity data sample in the indirect carbon footprint activity data sample, and calculate the indirect carbon footprint activity data based on the carbon emission factor sample to obtain the indirect-total carbon emission sample.

[0088] A370: Identify the indirect-energy-saving carbon footprint activity data sample of the indirect carbon footprint activity data sample, and the indirect-carbon emission reduction sample corresponding to the indirect-energy-saving carbon footprint activity data sample.

[0089] A380: Under the climate benefit index system, the indirect carbon emission reduction sample and the indirect carbon emission total sample are trained to obtain an indirect climate benefit measurement sub-model.

[0090] In one embodiment, firstly, indirect carbon footprint activity data samples of indirect carbon emission management units are identified. Multimodal data is extracted from these units: For example, at the textual feature level, activity descriptions are filtered to include raw material procurement / transportation and delivery; location information includes records of supplier factory areas / transportation routes, such as a 500km raw material transportation route for a project, and first-tier supplier capacity data; at the numerical feature level, parameters such as transportation fuel consumption (15 liters / 100km) and supplier electricity consumption (5000kWh / ton of raw material) are collected; at the categorical feature level, emission source type is labeled as indirect / supply chain, and industry code is defined as logistics / raw material production; at the spatiotemporal feature level, timestamps are extracted to indicate raw material procurement cycles, and supply chain levels are defined as first-tier / second-tier suppliers. Taking an electronics company as an example, an indirect carbon footprint dataset containing 800 samples is constructed, covering indirect links such as packaging material transportation and chip manufacturing.

[0091] Secondly, the specific steps for obtaining indirect carbon emission factor samples are detailed in A361-A362, and the total indirect carbon emissions are calculated. Unlike direct emission carbon emission factor samples, indirect carbon emission factor samples need to be obtained through conversion of direct emission carbon emission factor samples or from industry databases. For example, if the fuel consumption for raw material transportation is 15 liters / 100 kilometers, matching the direct emission carbon emission factor sample of 2.3 kg CO2 / liter, and considering a transportation distance of 500 kilometers, the emission at this stage is calculated to be 15 × 5 × 2.3 = 172.5 kg CO2. If the chip foundry supplier consumes 5000 kWh of electricity per ton of chip, through an indirect conversion module, the direct emission carbon emission factor sample of electricity production (0.8 kg CO2 / kWh) is converted into an indirect carbon emission factor sample, and the indirect emission per ton of chip is calculated to be 5000 × 0.8 = 4000 kg CO2. Summarizing the emissions from each indirect stage, the total annual indirect carbon emissions sample for this electronics company is obtained as 1200 tons of CO2, achieving the quantification from supply chain data to the total indirect emissions.

[0092] Next, identify indirect energy-saving carbon footprint activity data and indirect carbon emission reduction samples. Screen data on energy-saving retrofits in indirect processes: for example, after switching to electric trucks for raw material transportation, fuel consumption drops to 10 liters / 100 km, resulting in an annual emission reduction of (15-10)×5×2.3=57.5 kg CO2; a chip supplier upgrading to photovoltaic power supply reduces unit power consumption to 4500 kWh / ton, resulting in an annual emission reduction of (5000-4500)×0.8×10=4000 kg CO2, assuming an annual chip purchase of 10 tons. Label these data as indirect energy-saving carbon footprint samples, calculate emission reductions (before retrofit - after retrofit), forming indirect carbon emission reduction samples. For example, this company indirectly reduces emissions by 4057.5 kg annually, quantifying the climate contribution of energy conservation in the supply chain.

[0093] Finally, the model is trained under the climate benefit index system. The aforementioned four-dimensional indicators, including carbon emission reduction benefits and energy efficiency improvement benefits, are used: For example, carbon emission reduction benefits correspond to an indirect emission reduction of 4.0575 tons of CO2; in energy efficiency improvement benefits, the supplier's unit chip power consumption reduction = (5000-4500) / 5000 = 0.1, corresponding to an improvement coefficient of 0.1; the resource recycling benefit (e.g., increasing the packaging material recycling rate from 20% to 40%) has an impact weight of 0.2 × 1.5 (correction coefficient) = 0.3; the environmental synergy benefit (e.g., a supplier's nitrogen oxide emission reduction of 1 ton) is converted to a CO2 equivalent of 10 tons, with an improvement coefficient of 10 / 4.0575 × 0.7 = 1.725. These parameters are input into the neural network learning model, using the climate benefit scores of historical indirect emission projects (e.g., a supply chain project with a comprehensive score of 75 points) as labels, and iteratively trained for 200 rounds until the error is below 8%, thus constructing a sub-model for measuring indirect climate benefits. The model takes indirect carbon footprint data, emission factors, and four-dimensional index parameters as inputs and outputs a climate benefit score for indirect processes.

[0094] By identifying indirect emissions data from the supply chain, converting emissions factors, and combining them with multidimensional index training, the project achieved accurate quantification of indirect climate benefits, filling a gap in existing technologies for supply chain carbon emissions assessment.

[0095] Furthermore, step A360 in the method provided in this application embodiment includes:

[0096] A361: Obtain the direct carbon emission factor sample corresponding to the indirect carbon footprint activity data sample.

[0097] A362: Transform the direct carbon emission factor samples according to the indirect influence relationship to output indirect carbon emission factor samples.

[0098] Optionally, firstly, obtain the direct carbon emission factor samples corresponding to the indirect carbon footprint activity data samples. Taking the chip transportation of an electronics company as an example, its indirect carbon footprint data is 500 kilometers of road transportation with a fuel consumption of 15 liters per 100 kilometers. The corresponding direct carbon emission factor is 2.3 kg CO2 / liter from diesel combustion. The data comes from the industry carbon emission data guidelines. Through the indirect conversion module, the direct factors corresponding to this transportation activity are retrieved, and a mapping relationship between transportation fuel consumption and direct emissions is established.

[0099] Next, the direct carbon emission factor samples are converted according to the indirect impact relationships. Indirect impact factors are analyzed: for example, a 20% empty load rate (increasing energy consumption by 15%) and additional packaging material weight (increasing fuel consumption by 10%), determining a comprehensive conversion coefficient of 1.25 (1 + 15% + 10%). Multiplying the direct factor of 2.3 kg CO2 / L by the conversion coefficient yields the indirect carbon emission factor of 2.3 × 1.25 = 2.875 kg CO2 / L. This process needs to be based on historical transportation data statistics, such as an average empty load rate of 20% for road transportation in a certain region. The influence weights are trained using a random forest learning algorithm to ensure the accuracy of the conversion coefficient.

[0100] The final output is an indirect carbon emission factor sample. For example, the indirect factor for chip transportation in the above example is 2.875 kg CO2 / L, which effectively supports carbon asset accounting and green credit applications.

[0101] By acquiring direct carbon emission factors and converting them in conjunction with indirect influence relationships, the indirect carbon emission factors can be accurately calculated. This solves the problem of large deviations in the calculation of indirect emission factors in existing technologies and provides a reliable quantitative basis for the management of carbon footprint across the entire chain.

[0102] Furthermore, step A600 in the method provided in this application embodiment includes:

[0103] A610: Obtain feedback data from third-party calculations within the climate investment and financing platform.

[0104] A620: Retrain the direct-driven climate benefit calculation sub-model and the indirect-driven climate benefit calculation sub-model according to the calculated feedback data, and update the dual-driven climate benefit calculation model.

[0105] In this embodiment of the application, third-party measurement feedback data refers to measured data and evaluation results provided by an independent institution or entity outside the climate investment and financing platform for the purpose of verifying and optimizing the climate benefit measurement model.

[0106] In one embodiment, feedback data is first obtained from a third-party data interface of the climate investment and financing platform. This data includes: actual assessments of the project's climate benefits by financial institutions, such as a green credit project whose actual emission reductions, after a third-party audit, are 8% lower than the model's calculations; updates to industry emission benchmarks issued by regulatory authorities, such as the national updated carbon emission factors for the chemical industry; and on-site monitoring data from independent carbon verification agencies, such as a report on the difference between measured CO2 concentrations at a power plant chimney and model predictions. The climate investment and financing platform ensures the immutability of the data through blockchain smart contracts; for example, each feedback data entry is accompanied by a timestamp and a third-party electronic signature.

[0107] Secondly, the feedback data is standardized. Feedback data from different sources is mapped to the model input format. For example, the "actual emission reduction deviation rate" of financial institutions is converted into the "total carbon emission adjustment coefficient" required for model training, and industry benchmark updates are converted into "emission factor correction values". Taking a certain automobile manufacturing project as an example, third-party feedback indicated that its supply chain transportation emission factor was underestimated by 12%. After processing, a correction coefficient of 1.12 was obtained for subsequent training.

[0108] Then, iterative model training is initiated. Standardized feedback data is used as new training samples and input into the direct-driven and indirect-driven climate benefit measurement sub-models. Transfer learning techniques are employed to retain the general features already learned by the original model, such as the basic energy conversion algorithm, while focusing training only on the parts with larger biases (such as the indirect emission conversion coefficients of specific industries). For example, when multiple third-party feedback indicates that the indirect emission measurement bias of a certain type of logistics project exceeds 10%, the model automatically increases the training weight of this type of data to 30%.

[0109] Finally, cross-validation is used to ensure the effectiveness of the model update. The updated model is then used to backtest historical projects. If the historical calculation error drops below 5% after the update, the update is considered effective; otherwise, the training parameters are readjusted. The updated dual-drive model will be synchronized to the climate investment and financing platform for use in subsequent project calculations.

[0110] By constructing a closed-loop mechanism of feedback-training-validation and continuously optimizing the model using third-party dynamic data, the model has been able to adapt to technological upgrades and deviations in measured data, ensuring a dynamic match between climate benefit assessments and actual emission characteristics.

[0111] In summary, the climate benefit-driven project lifecycle carbon footprint management method provided in this application has the following technical effects:

[0112] This application obtains multiple project processes throughout the entire project lifecycle from an enterprise, separates these processes, establishes direct and indirect carbon emission management units for each process, synthesizes multiple direct and indirect carbon emission management units, and constructs a dual-drive climate benefit calculation model that includes direct and indirect climate benefit calculation sub-models. The sub-models are used to calculate direct and indirect calculation indicators, and the comprehensive calculation indicators are then sent to a climate investment and financing platform for storage. This accurately measures the carbon footprint throughout the project lifecycle, making the climate benefit assessment results more precise and reliable. It achieves the technical effect of finely quantifying the carbon footprint throughout the project lifecycle and accurately calculating climate benefits, providing comprehensive and accurate data support for project evaluation and carbon asset pricing in carbon emission trading.

[0113] Example 2, as Figure 2As shown, based on the same inventive concept as in Embodiment 1 above, this application provides a climate benefit-driven project lifecycle carbon footprint management system, the system comprising:

[0114] Project process acquisition module 1 is used to acquire multiple project processes throughout the entire lifecycle of an enterprise project.

[0115] Project process separation module 2 is used to separate the multiple project processes, establish a direct carbon emission management unit and an indirect carbon emission management unit for each project process, and synthesize multiple direct carbon emission management units and multiple indirect carbon emission management units for the multiple project processes.

[0116] Climate model construction module 3 is used to construct a dual-drive climate benefit calculation model, which includes a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model.

[0117] The calculation index acquisition module 4 is used to calculate multiple direct emission management units using the direct-driven climate benefit calculation sub-model and to calculate multiple indirect emission management units using the indirect-driven climate benefit calculation sub-model, so as to obtain direct calculation indexes and indirect calculation indexes.

[0118] The measurement index storage module 5 is used to calculate the comprehensive measurement index of the direct measurement index and the indirect measurement index, and send the comprehensive measurement index to the climate investment and financing platform for storage.

[0119] Furthermore, the project process separation module 2 is used to perform the following steps:

[0120] Establish direct carbon emission management units and indirect carbon emission management units for each project process. The direct carbon emission management units and the indirect carbon emission management units are connected to the carbon emission data acquisition module through a data interface to collect carbon footprint activity data of each carbon emission management unit. The carbon footprint activity data includes energy consumption monitoring data, material input and output records, production operation data, and environmental emission monitoring data.

[0121] Furthermore, the project process separation module 2 is used to perform the following steps:

[0122] The carbon emission data acquisition module includes a multimodal fusion classifier, which receives the carbon footprint activity data, classifies the data, and outputs at least one classification label, including direct emission labels and indirect emission labels; and routes the corresponding data to the direct carbon emission management unit or the indirect carbon emission management unit according to the at least one output classification label.

[0123] Furthermore, the project process separation module 2 is used to perform the following steps:

[0124] The multimodal fusion classifier includes an input layer, a feature fusion layer, a classification decision layer, and a label output layer. The input layer includes a text feature extractor, a numerical feature extractor, a category feature encoder, and a time series feature extractor. Historical carbon footprint activity data samples are collected, and text features, numerical features, category features, and spatiotemporal features are extracted using the input layer. These features are then input into the feature fusion layer to obtain fused feature samples. The classification decision layer is trained in multiple stages based on these fused feature samples until a multimodal fusion classifier with a preset accuracy is obtained. The trained multimodal fusion classifier is then deployed to the carbon emission data acquisition module.

[0125] Furthermore, the climate model building module 3 is used to perform the following steps:

[0126] Identify direct carbon footprint activity data samples of the direct carbon emission management unit; match direct carbon emission factor samples for each type of carbon footprint activity data sample in the direct carbon footprint activity data sample, and calculate the direct carbon footprint activity data based on the direct carbon emission factor samples to obtain a direct-total carbon emission sample; identify direct-energy-saving carbon footprint activity data samples of the direct carbon footprint activity data samples, and direct-carbon emission reduction samples corresponding to the direct-energy-saving carbon footprint activity data samples; define a climate benefit index system, and train the direct-carbon emission reduction sample and the direct-total carbon emission sample under the climate benefit index system to obtain a direct-driven climate benefit measurement sub-model.

[0127] Furthermore, the climate model building module 3 is used to perform the following steps:

[0128] The climate benefit index system includes carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits. The system calculates the direct emission reduction contribution value of the carbon emission reduction sample based on the carbon emission reduction benefits; calculates the energy efficiency improvement coefficient corresponding to the carbon emission reduction sample based on the energy efficiency improvement benefits; calculates the influence weight between the carbon emission reduction sample and the total carbon emission sample based on the resource recycling benefits; calculates the environmental synergy improvement coefficient of the carbon emission reduction sample based on the environmental synergy benefits; and uses the direct emission reduction contribution value, the energy efficiency improvement coefficient, the influence weight, and the synergy improvement coefficient as training parameters to construct a direct-driven climate benefit measurement sub-model.

[0129] Furthermore, the climate model building module 3 is used to perform the following steps:

[0130] Identify indirect carbon footprint activity data samples of the indirect carbon emission management unit; obtain indirect carbon emission factor samples for each type of carbon footprint activity data sample in the indirect carbon footprint activity data sample; calculate the indirect carbon footprint activity data based on the carbon emission factor samples to obtain an indirect-total carbon emission sample; identify indirect-energy-saving carbon footprint activity data samples of the indirect carbon footprint activity data sample, and indirect-carbon emission reduction samples corresponding to the indirect-energy-saving carbon footprint activity data samples; train the indirect-carbon emission reduction samples and the indirect-total carbon emission samples on climate benefits under the climate benefit index system to obtain an indirect-driven climate benefit measurement sub-model.

[0131] Furthermore, the climate model building module 3 is used to perform the following steps:

[0132] Obtain the direct carbon emission factor samples corresponding to the indirect carbon footprint activity data samples; transform the direct carbon emission factor samples according to the indirect influence relationship, and output the indirect carbon emission factor samples.

[0133] Furthermore, the calculation index storage module 5 is used to perform the following steps:

[0134] Obtain third-party calculation feedback data from the climate investment and financing platform; retrain the direct-driven climate benefit calculation sub-model and the indirect-driven climate benefit calculation sub-model according to the calculation feedback data, and update the dual-driven climate benefit calculation model.

[0135] The climate benefit-driven project lifecycle carbon footprint management system provided in this embodiment of the invention can execute the climate benefit-driven project lifecycle carbon footprint management method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0136] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0137] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A climate-benefit-driven approach to project lifecycle carbon footprint management, characterized in that, The method includes: Obtain multiple project processes throughout the entire lifecycle of an enterprise project; The multiple project processes are separated, and a direct carbon emission management unit and an indirect carbon emission management unit are established for each project process. The multiple direct carbon emission management units and multiple indirect carbon emission management units of the multiple project processes are then combined. A dual-drive climate benefit calculation model is constructed, which includes a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model. The direct-driven climate benefit calculation sub-model is used to calculate multiple direct emission management units, and the indirect-driven climate benefit calculation sub-model is used to calculate multiple indirect emission management units, to obtain direct and indirect calculation indicators. A comprehensive measurement index is calculated by combining the directly measured index and the indirectly measured index, and the comprehensive measurement index is sent to the climate investment and financing platform for storage.

2. The method as described in claim 1, characterized in that, Establish direct carbon emission management units and indirect carbon emission management units for each project process. The direct carbon emission management units and the indirect carbon emission management units are connected to the carbon emission data acquisition module through a data interface to collect carbon footprint activity data of each carbon emission management unit. The carbon footprint activity data includes energy consumption monitoring data, material input and output records, production operation data, and environmental emission monitoring data.

3. The method as described in claim 2, characterized in that, The carbon emission data acquisition module includes a multimodal fusion classifier, which is used to receive the carbon footprint activity data, classify the data, and output at least one classification label, which includes a direct emission label and an indirect emission label. The corresponding data is routed and written to the direct carbon emission management unit or the indirect carbon emission management unit based on at least one output classification label.

4. The method as described in claim 3, characterized in that, The multimodal fusion classifier includes an input layer, a feature fusion layer, a classification decision layer, and a label output layer. The input layer includes a text feature extractor, a numerical feature extractor, a category feature encoder, and a time series feature extractor. Historical carbon footprint activity data samples are collected, and text features, numerical features, category features and spatiotemporal features are extracted using the input layer. These features are then input into the feature fusion layer to obtain fused feature samples. Based on these fused feature samples, the classification decision layer is trained in multiple stages until a multimodal fusion classifier with a preset accuracy is obtained. The trained multimodal fusion classifier is deployed to the carbon emission data acquisition module.

5. The method as described in claim 1, characterized in that, Methods for constructing sub-models that directly drive climate benefits include: Identify a sample of direct carbon footprint activity data for the direct carbon emission management unit; The direct carbon emission factor sample is matched with the direct carbon footprint activity data sample of each type of carbon footprint activity data sample, and the direct carbon footprint activity data is calculated based on the direct carbon emission factor sample to obtain the direct-carbon emission total sample. Identify the direct-energy-saving carbon footprint activity data sample of the direct carbon footprint activity data sample, and the direct-carbon emission reduction sample corresponding to the direct-energy-saving carbon footprint activity data sample; A climate benefit index system is defined, and climate benefit training is performed on the direct carbon emission reduction sample and the total direct carbon emission sample under the climate benefit index system to obtain a direct-driven climate benefit measurement sub-model.

6. The method as described in claim 5, characterized in that, The climate benefit indicator system is used to train the carbon emission reduction sample and the total carbon emission sample on climate benefits. include: The climate benefit indicator system includes carbon emission reduction benefits, energy efficiency improvement benefits, resource recycling benefits, and environmental synergy benefits. Calculate the direct emission reduction contribution value of the carbon emission reduction sample based on the carbon emission reduction benefits. Calculate the energy efficiency improvement coefficient corresponding to the carbon emission reduction sample based on the energy efficiency improvement benefits; Calculate the influence weights of the carbon emission reduction sample and the total carbon emission sample based on the resource recycling benefits; Calculate the environmental synergistic improvement coefficient of the carbon emission reduction sample based on the aforementioned environmental synergistic benefits; The direct emission reduction contribution value, the energy efficiency improvement coefficient, the impact weight, and the synergistic improvement coefficient are used as training parameters to construct a sub-model for measuring climate benefits directly driven by climate.

7. The method as described in claim 5, characterized in that, Methods for constructing sub-models for indirectly driving climate benefits include: Identify data samples of the indirect carbon footprint activities of the indirect carbon emission management unit; Obtain indirect carbon emission factor samples for each type of carbon footprint activity data sample in the indirect carbon footprint activity data sample, and calculate the indirect carbon footprint activity data based on the carbon emission factor samples to obtain the indirect-total carbon emission sample. Identify the indirect-energy-saving carbon footprint activity data sample of the indirect carbon footprint activity data sample, and the indirect-carbon emission reduction sample corresponding to the indirect-energy-saving carbon footprint activity data sample; The climate benefit index system is used to train the indirect carbon emission reduction sample and the indirect carbon emission total sample to obtain the indirect climate benefit measurement sub-model.

8. The method as described in claim 7, characterized in that, The method for obtaining indirect carbon emission factor samples for each type of carbon footprint activity data sample from the indirect carbon footprint activity data sample via an indirect conversion module includes: Obtain the direct carbon emission factor samples corresponding to the indirect carbon footprint activity data samples; The direct carbon emission factor samples are transformed according to the indirect influence relationship to output indirect carbon emission factor samples.

9. The method as described in claim 1, characterized in that, Obtain feedback data from third-party calculations within the climate investment and financing platform; Based on the calculated feedback data, the direct-driven climate benefit calculation model and the indirect-driven climate benefit calculation model are retrained, and the dual-driven climate benefit calculation model is updated.

10. A climate-benefit-driven project lifecycle carbon footprint management system, characterized in that, For implementing the climate benefit-driven project lifecycle carbon footprint management method according to any one of claims 1-9, the system comprises: The project process acquisition module is used to acquire multiple project processes throughout the entire lifecycle of an enterprise project; The project process separation module is used to separate the multiple project processes, establish a direct carbon emission management unit and an indirect carbon emission management unit for each project process, and synthesize multiple direct carbon emission management units and multiple indirect carbon emission management units for the multiple project processes. The climate model construction module is used to construct a dual-drive climate benefit calculation model, which includes a direct-drive climate benefit calculation sub-model and an indirect-drive climate benefit calculation sub-model. The measurement index acquisition module is used to measure multiple direct emission management units using the direct-driven climate benefit measurement sub-model and to measure multiple indirect emission management units using the indirect-driven climate benefit measurement sub-model, so as to obtain direct measurement indexes and indirect measurement indexes. The measurement index storage module is used to calculate the comprehensive measurement index of the direct measurement index and the indirect measurement index, and send the comprehensive measurement index to the climate investment and financing platform for storage.

Images