Ai-driven dynamic life cycle assessment and scenario optimization method and system for product carbon footprint

Abstract

The application discloses an AI-driven dynamic life cycle assessment and scenario optimization method and system for product carbon footprint, obtains activity data, energy consumption data, transportation data and emission factor data of a target product in raw material acquisition, production and manufacturing, storage and logistics, use and recycling and disposal stages, constructs a life cycle process network after field alignment, time synchronization and unit conversion; identifies change events such as supplier change, process parameter adjustment, energy structure change, logistics path switching and emission factor update, determines the affected sub-network and performs local reconstruction; missing activity data, time-varying emission factors and implicit process emissions are predicted by using an artificial intelligence model, the life cycle inventory is updated after constraint correction, and the existing results of the unaffected part are combined to form a dynamic carbon footprint assessment result; further identify the key objects and generate the preferred scenario scheme. The method can improve the dynamic assessment efficiency and decision support capability.

Description

Technical Field

[0001] This invention relates to the field of product carbon footprint assessment technology, and in particular to an AI-driven dynamic life cycle assessment and scenario optimization method and system for product carbon footprint. Background Technology

[0002] Product carbon footprint refers to the total greenhouse gas emissions generated by a product throughout its entire life cycle, including raw material acquisition, manufacturing, warehousing and logistics, use, and recycling. Current product carbon footprint accounting is typically based on life cycle assessment methods, which calculate carbon emissions at each stage by establishing a life cycle inventory and combining it with emission factors.

[0003] Current technologies for assessing product carbon footprint mostly employ static modeling and full-scale accounting methods, which involve unified calculations based on fixed data and boundaries within a preset period. When suppliers, process parameters, energy structure, logistics routes, or emission factors change, it is often necessary to re-execute the full-process accounting, resulting in high computational costs, slow response times, and difficulty in timely reflecting changes in carbon emissions under actual production and operation conditions.

[0004] Furthermore, existing solutions typically focus on carbon footprint output, and are insufficient in supporting the propagation of local impacts caused by changing events, prediction of missing data, constraint correction, and scenario optimization based on assessment results. They are difficult to balance assessment timeliness, accounting accuracy, and optimization decision-making needs. Therefore, it is necessary to propose a product carbon footprint assessment solution that can perform local reconstruction, incremental accounting, and scenario optimization for dynamic events. Summary of the Invention

[0005] To address the problems of poor adaptability of static modeling, the need for full recalculation after the triggering of change events, insufficient timeliness of assessment, and weak support for optimization decisions in existing product carbon footprint assessments, this invention provides an AI-driven dynamic life cycle assessment and scenario optimization method and system for product carbon footprint.

[0006] In a first aspect, the present invention provides an AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint, comprising:

[0007] Acquire activity data, energy consumption data, transportation data, and emission factor data of the target product during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use, and recycling and disposal, and perform field alignment, time synchronization, and unit conversion;

[0008] A lifecycle process network is constructed based on the processed data. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0009] Identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and determine the affected sub-networks along the life cycle process network starting from the node corresponding to the event;

[0010] Based on the event, perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network;

[0011] Artificial intelligence models are used to predict missing activity data, time-varying emission factors, and hidden process emissions in affected subnetworks;

[0012] Based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints, the prediction results are truncated, backtracked, or re-estimated, and the lifecycle inventory is updated.

[0013] The carbon emissions of the affected subnetworks are recalculated, and the historical accounting results of the unaffected parts are retrieved and summarized according to node identifier, stage identifier or timestamp to obtain the dynamic carbon footprint assessment results.

[0014] Based on the dynamic carbon footprint assessment results, key process nodes or key connections with carbon emission contributions exceeding a preset threshold are identified. For these key process nodes or key connections, scenarios such as supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated and multi-objective optimization is performed. The optimal scenario scheme for updating the parameters of the life cycle process network is then output.

[0015] Secondly, it also provides an AI-driven dynamic lifecycle assessment and scenario optimization system for product carbon footprint, including:

[0016] The data acquisition and preprocessing module is used to acquire activity data, energy consumption data, transportation data and emission factor data of the target product in the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling and disposal, and to perform field alignment, time synchronization and unit conversion.

[0017] The process network construction module is used to construct a life cycle process network based on the processed data. The life cycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0018] The event identification module is used to identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and to determine the affected sub-network along the life cycle process network starting from the node corresponding to the event.

[0019] The dynamic reconfiguration module is used to perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network based on the event.

[0020] The prediction module is used to predict missing activity data, time-varying emission factors, and implicit process emissions in the affected subnetworks using artificial intelligence models;

[0021] The correction and update module is used to perform truncation correction, backtracking correction, or revaluation correction on the prediction results based on mass conservation, energy conservation, process flow, and equipment capacity constraints, and to update the lifecycle inventory.

[0022] The incremental accounting module is used to recalculate carbon emissions for the affected sub-networks, retrieve historical accounting results of the unaffected parts according to node identifier, stage identifier or timestamp, and summarize them to obtain dynamic carbon footprint assessment results.

[0023] The scenario optimization module is used to identify key process nodes or key connections with carbon emission contributions exceeding a preset threshold based on the dynamic carbon footprint assessment results. It generates supplier replacement, process parameter optimization, energy substitution, and logistics switching scenarios for the key process nodes or key connections and performs multi-objective optimization, outputting the preferred scenario scheme for updating the parameters of the life cycle process network.

[0024] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint as described above.

[0025] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements an AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint as described above.

[0026] Fifthly, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements an AI-driven dynamic lifecycle assessment and scenario optimization method for the carbon footprint of a product as described above.

[0027] Compared with existing technologies, this invention constructs a lifecycle process network to structurally correlate activity data, energy consumption data, transportation data, and emission factor data at each stage of the product's entire lifecycle. When events such as supplier changes, process adjustments, energy structure changes, logistics route switching, or emission factor updates are detected, dynamic reconstruction and incremental accounting are performed only on the affected sub-network, avoiding repetitive calculations throughout the entire process, thus significantly improving the efficiency and response speed of carbon footprint assessment. Simultaneously, artificial intelligence models are used to predict missing activity data, time-varying emission factors, and implicit process emissions, and corrections are made in conjunction with mass conservation, energy conservation, process flow, and equipment capacity constraints, contributing to the completeness and accuracy of dynamic assessment results. Furthermore, this invention can identify key process nodes or key connection edges based on dynamic carbon footprint assessment results, and further generate and optimize scenario solutions such as supplier replacement, process parameter optimization, energy substitution, and logistics switching, thereby balancing multiple objectives such as carbon emission reduction, cost control, delivery cycle, and quality risk, enhancing the intelligence level and application value of product carbon management. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint provided in this embodiment of the invention.

[0029] Figure 2 This is a schematic diagram of the structure of the AI-driven dynamic life cycle assessment and scenario optimization system for product carbon footprint provided in an embodiment of the present invention;

[0030] Figure 3 An embodiment diagram of the electronic device provided in this invention;

[0031] Figure 4 An embodiment diagram of a computer-readable medium provided for embodiments of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0034] In the description of this invention, the term "such as" is used to mean "used as an example, illustration, or description." Any embodiment described "such as" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to implement and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be implemented without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0035] See Figure 1 , Figure 1 This is a flowchart illustrating the AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint provided by the present invention. In this embodiment of the invention, the executing entity of the AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint is the AI-driven dynamic lifecycle assessment and scenario optimization system for product carbon footprint. Therefore, the AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint includes...

[0036] Step 10: Obtain activity data, energy consumption data, transportation data, and emission factor data of the target product during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use, and recycling and disposal, and perform field alignment, time synchronization, and unit conversion.

[0037] This step provides a unified data foundation for subsequent lifecycle process network construction, change event identification, affected subnetwork determination, local reconstruction, dynamic accounting, and scenario optimization.

[0038] Specifically, activity data, energy consumption data, transportation data, and emission factor data of the target product are acquired during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling and disposal. Field alignment, time synchronization and unit conversion are performed on the acquired data to form a preprocessed data set that can be used for unified accounting.

[0039] Optionally, the above data can be categorized by stage, process, or object so that it can be subsequently attached to the corresponding process node or connection edge.

[0040] Furthermore, the preprocessed data can maintain consistency in time scale, unit system, and field structure, providing a directly accessible data foundation for subsequent lifecycle process network modeling.

[0041] Specifically, see the descriptions in steps 101-104.

[0042] Step 20: Construct a lifecycle process network based on the processed data. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges, and transportation connection edges between nodes.

[0043] This step is used to structurally express the unit processes and their relationships throughout the product lifecycle as a computable and updatable network model.

[0044] Specifically, a lifecycle process network is constructed based on the preprocessed data obtained in step 10. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges, and transportation connection edges between nodes; wherein, each process node corresponds to a lifecycle unit process, and each connection edge is used to characterize the transmission relationship between adjacent processes.

[0045] Optionally, process nodes can be associated with activity levels, emission factors, time markers, and stage markers to characterize the status of the node within a specific accounting period.

[0046] Furthermore, by constructing a lifecycle process network, the originally scattered data can be integrated into a unified network model, which facilitates subsequent local updates and dynamic accounting around change events.

[0047] Specifically, see the descriptions in steps 201-203.

[0048] Step 30: Identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and determine the affected sub-networks along the life cycle process network, starting from the node corresponding to the event.

[0049] This step is used to identify events that cause changes in carbon emissions on a life cycle process network basis and to determine the scope of the impact of such events within the network.

[0050] Specifically, the system identifies change events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and maps the identified change events to the corresponding process nodes. Starting from these process nodes, the system determines the affected sub-networks along the lifecycle process network.

[0051] Optionally, the scope of influence can be determined by considering the dependencies between processes, material transfer relationships, energy transfer relationships, or transportation relationships.

[0052] Furthermore, by extracting the affected sub-networks related to the change events from the complete lifecycle process network, the processing scope can be limited for subsequent local reconstruction and incremental accounting.

[0053] Specifically, see the descriptions in steps 301-303.

[0054] Step 40: Perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network based on the event.

[0055] This step is used to partially update the lifecycle process network based on changing events, without having to rebuild the entire lifecycle model.

[0056] Specifically, based on the change events, the lifecycle process network performs process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update.

[0057] Optionally, when the supplier changes, the process nodes of the corresponding raw material acquisition stage can be replaced; when the logistics route changes, the relevant transportation connection edges can be adjusted; when the energy structure changes, the relevant energy supply process nodes and their emission factors can be updated; when the accounting scope changes, the system boundary can be adjusted.

[0058] Furthermore, the dynamically reconstructed lifecycle process network still maintains the relationships between each stage, so that prediction, correction and accounting can continue to be carried out on the updated network.

[0059] Specifically, see the descriptions in steps 401-403.

[0060] Step 50: Use artificial intelligence models to predict missing activity data, time-varying emission factors, and implicit process emissions in the affected subnetworks.

[0061] This step is used to predict accounting parameters in the affected subnetwork that were not fully collected or that changed over time.

[0062] Specifically, artificial intelligence models are used to predict missing activity data, time-varying emission factors, and implicit process emissions in the affected subnetworks.

[0063] Optionally, the input to the artificial intelligence model may include historical activity data, real-time equipment parameters, process parameters, transportation distances, and historical emission factor sequences, and the output may be the prediction results required for subsequent carbon emission accounting.

[0064] Furthermore, the implicit process emissions refer to process emissions that are not directly monitored or collected, but can be inferred from activity status, process parameters, or equipment operating status. This step can improve the continuity and integrity of the data from the affected sub-networks.

[0065] Specifically, see the descriptions in steps 501-503.

[0066] Step 60: Based on the conservation of mass, conservation of energy, process flow and equipment capacity constraints, perform truncation correction, backtracking correction or re-estimation correction on the prediction results, and update the life cycle inventory.

[0067] This step is used to verify the prediction results against the actual process conditions and accounting logic in order to improve the rationality of subsequent dynamic accounting.

[0068] Specifically, based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints, the prediction results are subject to truncation correction, backward correction, or reassessment correction, and the lifecycle inventory is updated.

[0069] Optionally, when the prediction results exceed the equipment's capabilities or do not meet existing process logic, the corresponding results can be corrected; when the lack of key inputs leads to insufficient prediction basis, the relevant parameters can be re-estimated.

[0070] Furthermore, the corrected activity levels, emission factors, implicit process emissions, and corresponding timestamps will be written into the process nodes or connecting edges of the affected subnetwork for subsequent carbon emission recalculation.

[0071] Specifically, see the descriptions in steps 601-603.

[0072] Step 70: Recalculate the carbon emissions of the affected subnetworks, retrieve the historical accounting results of the unaffected parts according to node identifier, stage identifier or timestamp, and summarize them to obtain the dynamic carbon footprint assessment results.

[0073] This step is used to generate a dynamic carbon footprint assessment result for the current accounting cycle based on partial updates.

[0074] Specifically, carbon emissions are recalculated for the affected subnetworks, and existing accounting results for the unaffected parts are retrieved according to node identifiers, stage identifiers, or timestamps; the two are then combined to obtain dynamic carbon footprint assessment results.

[0075] Optionally, affected subnetworks are recalculated using the updated lifecycle inventory, while unaffected parts directly reuse existing results corresponding to the current accounting period.

[0076] Furthermore, by combining local recalculation with the reuse of existing results, the amount of redundant calculations can be reduced and the efficiency of dynamic evaluation can be improved.

[0077] Specifically, see the descriptions in steps 701-703.

[0078] Step 80: Based on the dynamic carbon footprint assessment results, identify key process nodes or key connection edges whose carbon emission contribution exceeds a preset threshold, generate supplier replacement, process parameter optimization, energy substitution and logistics switching scenarios for the key process nodes or key connection edges, perform multi-objective optimization, and output the preferred scenario scheme for updating the parameters of the life cycle process network.

[0079] This step is used to further transform the dynamic carbon footprint assessment results into optimization decision-making results.

[0080] Specifically, based on the dynamic carbon footprint assessment results, key process nodes or key connections with carbon emission contributions exceeding a preset threshold are identified. For these key process nodes or key connections, scenarios such as supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated. The generated scenarios are then optimized using multiple objectives, and the optimal scenario schemes for updating the parameters of the life cycle process network are output.

[0081] Alternatively, key targets can be determined based on the proportion of their carbon emissions to the total carbon footprint.

[0082] Furthermore, the preferred scenario solution can be fed back to the lifecycle process network to continuously update network parameters and support subsequent dynamic evaluation.

[0083] Specifically, see the descriptions in steps 801-804.

[0084] This invention constructs a lifecycle process network to uniformly link activity data, energy consumption data, transportation data, and emission factor data throughout the entire lifecycle of a target product. When events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, or emission factor updates occur, dynamic reconstruction and partial recalculation are performed only on the affected subnetworks. This helps reduce the computational burden caused by repeated accounting throughout the entire process and improves the response efficiency of carbon footprint assessment.

[0085] Meanwhile, this invention utilizes artificial intelligence models to predict missing activity data, time-varying emission factors, and implicit process emissions, and combines these with mass conservation, energy conservation, process flow, and equipment capacity constraints for correction, which helps to enhance the completeness and rationality of dynamic evaluation results.

[0086] Furthermore, this invention can identify key process nodes or key connection edges based on dynamic carbon footprint assessment results, and further generate scenarios such as supplier replacement, process optimization, energy substitution and logistics switching for multi-objective optimization, thereby taking into account factors such as carbon emissions, cost, delivery cycle and quality risks, and providing support for product carbon management and low-carbon decision-making.

[0087] In one embodiment, steps 101-104 are described as follows:

[0088] Step 101: Obtain raw data for each stage of the target product's entire lifecycle.

[0089] In step 101, raw data of the target product are obtained during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling.

[0090] The raw data includes at least activity data, energy consumption data, transportation data, and emission factor data. Activity data characterizes the input, output, or operational activity status of each life cycle unit process; energy consumption data characterizes the energy consumption status of each process or equipment; transportation data characterizes transportation routes, modes of transportation, transportation distances, or transportation operational status; and emission factor data characterizes the emission conversion basis for unit activity volume, unit energy consumption, unit transportation volume, or corresponding processes.

[0091] Optionally, the aforementioned raw data may come from the company's internal business records, process records, logistics records, usage records, or disposal records, or from an emission factor database.

[0092] Furthermore, data at different stages may differ in recording granularity, update frequency, and expression format. Therefore, when acquiring data, it can be categorized and stored according to the lifecycle stage so that it can be preprocessed and networked in stages later.

[0093] The output of step 101 is a raw data set categorized by stage, which provides the input basis for subsequent field alignment, time synchronization, and unit conversion.

[0094] Step 102: Perform field alignment on the original data.

[0095] In step 102, field alignment is performed on raw data from different sources and with different structures to unify the way the same accounting object or the same accounting element is expressed.

[0096] Specifically, fields that can characterize the same process object are mapped to a unified set of fields, so that activity volume, energy consumption, transportation volume, emission factor, stage identifier, object identifier, and time identifier have consistent meaning in subsequent processing.

[0097] Furthermore, after field alignment, standardized data records are formed. Each standardized data record includes at least an object identifier field, a stage identifier field, a time field, a data type field, and a numeric field. The time field is used to record the timestamp corresponding to the data, and the time synchronization is used to align data with different timestamps to a unified accounting period, thereby providing a foundation for subsequent time synchronization, unit conversion, and network connection.

[0098] The time field is used to record the timestamps corresponding to the data, and the time synchronization is used to align data with different timestamps to a unified accounting period.

[0099] Optionally, for fields with the same meaning but different names, a field mapping table can be pre-established; for the same field with different encodings in different data sources, it can be converted using a unified encoding rule.

[0100] The output of step 102 is a standardized set of data records with a unified field structure, which enables subsequent data from different sources to be merged under the same processing logic.

[0101] Step 103: Synchronize the time and convert the units of the standardized data records.

[0102] In step 103, the standardized data records obtained in step 102 are synchronized with time and converted to units.

[0103] Time synchronization is used to adjust data from different sampling periods, different recording times, or different statistical periods to a unified accounting period; unit conversion is used to convert data from different units of measurement into a unified accounting unit to ensure comparability in subsequent calculations.

[0104] Optionally, data with different time granularities can be aggregated, split, or aligned according to a predetermined accounting cycle; data with different units can be converted to a unified dimensional system according to a preset conversion relationship.

[0105] Furthermore, the data after time synchronization and unit conversion can reflect the status of different life cycle objects within the same accounting window, avoiding deviations in subsequent network modeling and carbon emission accounting caused by inconsistent recording periods or units.

[0106] The output of step 103 is preprocessed data with consistent time reference and unit system.

[0107] Step 104: Establish the mapping relationship between preprocessed data and lifecycle objects.

[0108] In step 104, the preprocessed data obtained in step 103 is mapped to process nodes or connection edges in the subsequent lifecycle process network.

[0109] Specifically, activity data, energy consumption data, and emission factor data related to the unit process are mapped to the corresponding process nodes, and transportation-related data are mapped to the corresponding transportation connection edges. These mappings can be established based on object identifiers, stage identifiers, and timestamps.

[0110] Step 104 allows various types of data to be associated with specific lifecycle objects, thus providing a foundation for the construction of the lifecycle process network in step 20.

[0111] Furthermore, the mapped data is no longer an isolated data record, but becomes attribute information that can be directly accessed by process nodes or connecting edges. Subsequently, in the processes of change event identification, local reconstruction, and dynamic accounting, the corresponding objects can be quickly located and their parameters updated based on this mapping relationship.

[0112] The output of step 104 is the connection result between the data and the lifecycle object.

[0113] In one embodiment, steps 201-203 are described as follows:

[0114] Step 201: Define the basic objects in the lifecycle process network.

[0115] In step 201, the basic objects in the lifecycle process network are defined based on the data connection results of step 104, including process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0116] Each process node corresponds to a life cycle unit process, and each connecting edge is used to represent the flow relationship between adjacent processes.

[0117] Each unit process in the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling can be abstracted as a process node; the material transfer, energy supply and transportation relationships between processes can be abstracted as different types of connection edges.

[0118] Furthermore, by abstracting lifecycle objects into nodes and edges, the relationships of the entire lifecycle process can be transformed from scattered records into a network model with structural relationships, providing a foundation for subsequent event propagation and local updates.

[0119] The output of step 201 is a collection of network objects.

[0120] Step 202: Configure attribute information for the network object.

[0121] In step 202, attribute information is configured for the process nodes and connection edges defined in step 201.

[0122] For process nodes, activity levels, emission factors, time markers, and stage markers can be configured; for connection edges, corresponding material flow rates, energy flow rates, transport volumes, or associated time information can be configured. The stage marker indicates the lifecycle stage of the node, and the time marker indicates the accounting cycle to which the object belongs.

[0123] Furthermore, time stamps can be used to match existing accounting results for unaffected portions during subsequent dynamic accounting; stage stamps can be used to output carbon emission results by stage.

[0124] Optionally, process nodes can also be associated with their respective process type information to quickly identify affected objects when a change event occurs.

[0125] The output of step 202 is a set of network objects with attribute information, which enables the life cycle process network to not only have structural expression capabilities, but also state carrying capabilities.

[0126] Step 203: Form a life cycle process network.

[0127] In step 203, the network objects and their attributes obtained in steps 201 and 202 are organized to form a lifecycle process network covering the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling.

[0128] This network can express the relationships between processes throughout the entire product lifecycle, and it can also carry the state parameters of each process within a specific accounting cycle.

[0129] Furthermore, the resulting lifecycle process network can serve as a unified data organization carrier, enabling data from different stages, sources, and types to establish relationships within the same network structure.

[0130] Therefore, subsequent change events no longer correspond to isolated data records, but rather to specific process nodes or connection edges in the network.

[0131] The life cycle process network obtained in step 203 serves as a unified basic model for subsequent event identification, affected subnetwork extraction, network dynamic reconstruction, prediction correction, carbon emission recalculation, and scenario optimization.

[0132] In one embodiment, steps 301-303 are described as follows:

[0133] Step 301: Identify change events and map them to the corresponding process nodes.

[0134] In step 301, events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates are identified based on changes in input data.

[0135] After a change event is identified, the change event is mapped to the corresponding process node, which is then used as the event-corresponding node.

[0136] Specifically, for change events that occur directly at the process node level, the corresponding process node can be directly identified; for change events related to connection edges, they can be mapped to the process node associated with that connection edge.

[0137] Furthermore, by uniformly mapping different types of change events to process nodes, subsequent impact scope identification can be carried out with a consistent object entry point, avoiding the fragmentation of processing flow due to different event carriers.

[0138] The output of step 301 is the event type and the corresponding node of the event.

[0139] Step 302: Identify the impact relationships of events along the lifecycle process network.

[0140] In step 302, starting from the node corresponding to the event determined in step 301, relevant process nodes and connecting edges that are dependent on the event are identified along the life cycle process network to determine the possible scope of the impact propagation caused by the change event.

[0141] The dependencies include material input / output dependencies, energy supply dependencies, or transportation path dependencies. During identification, the process relationships represented by material flow connection edges, energy flow connection edges, and transportation connection edges can be combined to determine objects directly or indirectly related to the corresponding node of the event.

[0142] Optionally, to characterize the degree to which different objects are affected by the changing event, an impact metric can be constructed:

[0143]

[0144] in, Indicates the first The impact metric for an object;

[0145] This represents the change in the object's activity relative to the baseline state;

[0146] This represents the baseline activity level of the object;

[0147] This indicates the change in the carbon emissions of the object relative to a baseline state;

[0148] This represents the baseline carbon emissions of the object;

[0149] and These are the weighting coefficients;

[0150] A positive number is set to avoid a denominator of zero.

[0151] The impact metric is used to help determine whether an object is included in the affected scope.

[0152] Furthermore, based on this impact metric, process nodes and connecting edges that are substantially related to the changing events can be identified more effectively.

[0153] Step 303: Identify the affected sub-networks.

[0154] In step 303, the affected sub-networks related to the change event are determined based on the relevant process nodes and connection edges identified in step 302.

[0155] Optionally, the scope of the affected subnetwork can be defined based on the level of propagation, the degree of change in activity, or the degree of change in carbon emissions, so that the subnetwork corresponds to the change event.

[0156] Step 303 allows the local impact range corresponding to the current change event to be extracted from the complete lifecycle process network, providing an object basis for the local reconstruction, local prediction, local correction, and local recalculation in subsequent steps 40 to 70.

[0157] Furthermore, identifying the affected subnetworks allows subsequent processing to focus on the changed parts, while the unaffected parts can remain in their existing state, thus providing a prerequisite for incremental dynamic evaluation.

[0158] In one embodiment, steps 401-403 are described as follows:

[0159] Step 401: Determine the network reconstruction method based on the change events.

[0160] In step 401, the corresponding network reconstruction method is determined based on the type of change event identified in step 301, including process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update.

[0161] Among them, supplier changes can correspond to process node replacements, logistics route switching can correspond to connection edge adjustments, energy structure changes can correspond to updates of relevant energy supply process nodes and their emission factors, and accounting scope changes can correspond to system boundary adjustments.

[0162] Among them, system boundary adjustment refers to updating the process nodes and connection edge ranges involved in the current product carbon footprint accounting.

[0163] Furthermore, by establishing a correspondence between change event types and reconstruction methods, event identification results can be transformed into executable network update operations.

[0164] The output of step 401 is the local reconstruction instruction corresponding to the change event.

[0165] Step 402: Perform local reconstruction on the affected subnetwork.

[0166] In step 402, based on the reconstruction method determined in step 401, only the affected subnetworks obtained in step 303 are partially reconstructed, while the unaffected parts remain unchanged.

[0167] When a process node is replaced, the original process's association in the network is transferred to the replaced process node; when a connection edge is adjusted, the transportation, material, or energy relationships between related nodes are updated; when the system boundary is adjusted, the scope of nodes and connection edges included in the current calculation is updated; when the emission factor is updated, the new emission factor is written to the corresponding process node.

[0168] Local reconstruction can reflect the impact of changing events on local structures and parameters while maintaining the overall stability of the lifecycle process network.

[0169] Furthermore, this partial reconstruction approach avoids remodeling the entire lifecycle process network, allowing subsequent accounting to be carried out only around the affected objects, thus improving the processing efficiency of the dynamic evaluation process.

[0170] Step 403: Verify the consistency of the reconstructed network.

[0171] In step 403, the affected sub-networks after reconstruction in step 402 are subjected to consistency verification to ensure that the reconstructed network still meets the structural association and attribute integrity requirements for subsequent calculations.

[0172] The consistency check includes at least: whether the connection relationship between process nodes is complete, whether the stage identifier is valid, whether the time identifier is traceable, and whether the accounting object is within the current system boundary.

[0173] Furthermore, if the consistency check finds broken node relationships, missing attributes, or objects exceeding the current system boundary, the process can return to the aforementioned partial reconstruction step to correct the corresponding nodes or connecting edges.

[0174] The consistency verification result obtained in step 403 is used to confirm that the reconstructed life cycle process network can continue to enter the prediction and accounting process, thereby ensuring that the subsequent dynamic evaluation is based on a structurally effective network.

[0175] In one embodiment, steps 501-503 are described as follows:

[0176] Step 501: Construct the input for the artificial intelligence model.

[0177] In step 501, input data for the artificial intelligence model is constructed for the relevant process nodes and connection edges in the affected sub-network.

[0178] The input data may include historical activity data, real-time equipment parameters, process parameters, transportation distance, and historical emission factor sequences. The historical activity data reflects the operating status of the object within the existing accounting cycle, the real-time equipment parameters and process parameters reflect the current process status, the transportation distance reflects the logistics status, and the historical emission factor sequences reflect the time-varying characteristics of emission factors.

[0179] Furthermore, the input data can be organized according to process nodes or connection edges, so that the model output can correspond one-to-one with specific lifecycle objects. In this way, after the model completes the prediction, the results can be directly attached to the corresponding objects in the affected sub-network.

[0180] The output of step 501 is the set of model input samples corresponding to the affected sub-network, providing a unified input basis for subsequent prediction steps.

[0181] Step 502: Predict missing activity data, time-varying emission factors, and implicit process emissions.

[0182] In step 502, the model input sample constructed in step 501 is input into the artificial intelligence model to obtain the prediction results of missing activity data, time-varying emission factors, and implicit process emissions.

[0183] For any process node Its predicted emissions can be expressed as:

[0184]

[0185] in, Represents a node The predicted emissions;

[0186] This represents the predicted activity level;

[0187] This represents the predicted time-varying emission factor;

[0188] This indicates the emissions from the predicted implicit processes.

[0189] The predicted emissions are used for subsequent constraint revisions and carbon emission recalculations.

[0190] Furthermore, step 502 is not limited to directly outputting the final carbon emission result, but more importantly, it outputs intermediate parameters for subsequent verification and correction, so that the affected subnetwork after dynamic reconstruction has a parameter basis for continued calculation.

[0191] Step 503: Form a set of prediction results for the affected subnetworks.

[0192] In step 503, the prediction results obtained in step 502 are aggregated according to process nodes or connection edges to form a set of prediction results for the affected subnetwork.

[0193] The set of prediction results includes at least object identifiers, stage identifiers, time identifiers, predicted activity levels, predicted emission factors, and predicted implicit process emissions.

[0194] Furthermore, by objectifying and aggregating the prediction results, subsequent constraint corrections can be made to process a set of prediction parameters that match the structure of the affected sub-network, rather than focusing on scattered prediction values.

[0195] The set of prediction results obtained in step 503 will be used as input for the constraint correction and lifecycle list update in step 60, thereby achieving a smooth connection from prediction results to accounting parameters.

[0196] In one embodiment, steps 601-603 are described as follows:

[0197] Step 601: Verify the prediction results based on the constraints.

[0198] In step 601, the set of prediction results obtained in step 503 is verified based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints.

[0199] Among them, quality conservation is used to verify the relationship between material input and output, energy conservation is used to verify the relationship between energy input and consumption, process flow is used to verify the logical relationship between processes, and equipment capacity constraint is used to verify whether the amount of activity exceeds the equipment's processing capacity.

[0200] Step 601 can identify objects in the prediction results that are out of bounds, unbalanced, or do not conform to the process logic.

[0201] Furthermore, after this verification step, it is possible to distinguish which prediction results can be directly used for subsequent calculations and which prediction results need to be corrected, thereby avoiding unreasonable parameters from being directly included in the carbon emission calculation process.

[0202] Step 602: Perform truncation correction, backtracking correction, or revaluation correction.

[0203] In step 602, the prediction results are corrected according to the verification results of step 601.

[0204] When a prediction result exceeds the equipment's capacity or the allowable range, a truncation correction is performed on the prediction result; when the prediction result does not satisfy the mass conservation or energy conservation relationship, a backtracking correction is performed on the relevant process nodes; when insufficient key inputs lead to instability in the prediction basis, a revaluation correction is performed on the corresponding object.

[0205] Among them, truncation correction is used to limit abnormal prediction values ​​to a reasonable range, backtracking correction is used to correct the prediction results of related objects based on the relationship between adjacent processes, and revaluation correction is used to re-form the prediction results based on the available inputs.

[0206] Furthermore, different correction methods correspond to different types of anomaly sources, thereby ensuring that the correction process remains consistent with actual process conditions and accounting logic.

[0207] The output of step 602 is the corrected activity level, emission factor, and implicit process emissions.

[0208] Step 603: Update the lifecycle inventory.

[0209] In step 603, the correction results obtained in step 602 are written into the process nodes and connection edges corresponding to the affected subnetwork, and the lifecycle list is updated.

[0210] The updated information should include at least the activity level, emission factors, implicit process emissions, and their corresponding timestamps.

[0211] Furthermore, the updated lifecycle inventory reflects the latest measurable status of the affected subnetworks under the influence of the change events. In subsequent step 70, the carbon emissions of the affected subnetworks can be recalculated directly based on this lifecycle inventory without having to return to the original data layer for processing.

[0212] The updated results from step 603 constitute the direct accounting input for the dynamic carbon footprint assessment.

[0213] In one embodiment, steps 701-703 are described as follows:

[0214] Step 701: Recalculate the carbon emissions of the affected subnetwork.

[0215] In step 701, carbon emissions are recalculated for process nodes and connection edges in the affected subnetwork based on the updated lifecycle list from step 603.

[0216] For any process node Its recalculated carbon emissions can be expressed as:

[0217]

[0218] in, Represents a node Carbon emissions;

[0219] This indicates the revised activity level;

[0220] This indicates the corrected emission factor;

[0221] This indicates the revised implicit process emissions.

[0222] This result is used to generate local accounting results for the affected subnetwork.

[0223] Furthermore, the recalculation objects in step 701 are limited to process nodes and connection edges in the affected sub-network. Therefore, compared to recalculating the entire lifecycle process network, the number of objects that are repeatedly calculated can be reduced, and this calculation can focus more on the changed parts.

[0224] Step 702: Retrieve the existing accounting results from the unaffected portion.

[0225] In step 702, based on the node identifier, stage identifier, and timestamp, the existing accounting results corresponding to the current accounting cycle for the unaffected part are retrieved, or the most recent valid accounting results corresponding to the current accounting cycle are retrieved.

[0226] Since this part was not affected by this change, it does not need to be recalculated.

[0227] Step 702 allows the existing results of the unaffected portion to be directly incorporated into the results of this dynamic accounting.

[0228] Furthermore, this step enables dynamic carbon footprint assessment to include not only partial recalculation results but also existing results from the stable portion, thereby reducing overall accounting overhead while ensuring the integrity of the results.

[0229] Step 703: Summarize the dynamic carbon footprint assessment results.

[0230] In step 703, the local accounting results of the affected subnetwork obtained in step 701 are combined with the existing accounting results of the unaffected part obtained in step 702 to form a dynamic carbon footprint assessment result.

[0231] Its summary form can be expressed as:

[0232]

[0233] in, This indicates the results of a dynamic carbon footprint assessment;

[0234] Represents the set of objects in the affected subnetwork;

[0235] Represents a set medium object The recalculated carbon emissions;

[0236] This represents a collection of objects that are unaffected.

[0237] Represents a set medium object The existing accounting results.

[0238] The results obtained in step 703 are used in step 80 to identify key process nodes or key connection edges, and serve as the evaluation basis for scenario optimization.

[0239] Furthermore, this aggregation method allows for the generation of complete product carbon footprint results even under conditions of localized changes.

[0240] In one embodiment, steps 801-804 are described as follows:

[0241] Step 801: Identify key process nodes or key connection edges.

[0242] In step 801, the carbon emission contribution of each process node and connection edge is calculated based on the dynamic carbon footprint assessment results obtained in step 703.

[0243] Process nodes or connecting edges whose carbon emission contribution exceeds a preset threshold are identified as critical process nodes or critical connecting edges.

[0244] Step 801 allows subsequent scenario optimization to focus on objects that have a more significant impact on the total carbon footprint.

[0245] Furthermore, identifying key targets can prevent the generation and evaluation of too many solutions for targets with low contribution, thus making the subsequent optimization process more targeted.

[0246] Step 802: Generate candidate optimization scenarios.

[0247] In step 802, for the key process nodes or key connection edges identified in step 801, scenarios for supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated.

[0248] Different candidate scenarios correspond to different network parameter update methods. Among them, the supplier replacement scenario is used to update the process parameters related to the raw material acquisition stage, the process parameter optimization scenario is used to update the process parameters related to the production and manufacturing stage, the energy substitution scenario is used to update the process parameters related to energy supply, and the logistics switching scenario is used to update the transportation connection edge parameters.

[0249] Furthermore, each candidate optimization scenario can correspond to a set of process node parameters or connection edge parameters to be updated, thereby enabling different optimization schemes to be compared within the same lifecycle process network framework.

[0250] The output of step 802 is a set of candidate optimization scenarios.

[0251] Step 803: Perform multi-objective optimization on the candidate scenarios.

[0252] In step 803, a multi-objective optimization evaluation is performed on the candidate optimization scenarios generated in step 802.

[0253] Optionally, multi-objective optimization can comprehensively consider carbon emissions, production costs, delivery cycles, and quality risks to determine the preferred solution that meets the preset optimization objectives from multiple candidate scenarios.

[0254] Furthermore, the aforementioned carbon emissions can be obtained from the dynamic accounting results under the scenario corresponding to step 703, and production costs, delivery cycles, and quality risks can be calculated or assessed based on the supply, process, and logistics parameters corresponding to the candidate scenarios.

[0255] By comprehensively evaluating multiple objectives, the final selected scenario can not only focus on carbon emission reduction effects, but also take into account implementation feasibility and application value.

[0256] The output of step 803 is the scenario evaluation result.

[0257] Step 804: Output the preferred scenario and update the network parameters.

[0258] In step 804, based on the scenario evaluation results obtained in step 803, the preferred scenario scheme is determined, and its corresponding process parameters or connection edge parameters are fed back to update the life cycle process network.

[0259] This method allows optimization results to be written back to the lifecycle process network after a dynamic evaluation is completed, providing a foundation for subsequent continuous evaluation and optimization.

[0260] Furthermore, after the optimal scenario solution is written back, the relevant parameters in the lifecycle process network are updated synchronously, so that when new change events occur, dynamic evaluation can continue to be carried out on the basis of the updated network.

[0261] Thus, this invention achieves a closed-loop processing procedure from event identification, local reconstruction, dynamic evaluation to scenario optimization.

[0262] Optionally, refer to Figure 2 , Figure 2This is a schematic diagram of the structure of the AI-driven dynamic life cycle assessment and scenario optimization system for product carbon footprint provided by the present invention. The AI-driven dynamic life cycle assessment and scenario optimization system for product carbon footprint includes...

[0263] The data acquisition and preprocessing module 210 is used to acquire activity data, energy consumption data, transportation data and emission factor data of the target product in the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling and disposal, and to perform field alignment, time synchronization and unit conversion.

[0264] The process network construction module 220 is used to construct a life cycle process network based on the processed data. The life cycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0265] Event identification module 230 is used to identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and to determine the affected sub-network along the life cycle process network starting from the node corresponding to the event.

[0266] The dynamic reconfiguration module 240 is used to perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network according to the event.

[0267] Prediction module 250 is used to predict missing activity data, time-varying emission factors and implicit process emissions in the affected subnetwork using an artificial intelligence model;

[0268] The correction and update module 260 is used to perform truncation correction, backtracking correction, or revaluation correction on the prediction results based on mass conservation, energy conservation, process flow, and equipment capacity constraints, and to update the lifecycle list.

[0269] The incremental accounting module 270 is used to recalculate the carbon emissions of the affected sub-networks, retrieve the historical accounting results of the unaffected parts according to the node identifier, stage identifier or timestamp and summarize them to obtain the dynamic carbon footprint assessment results.

[0270] The scenario optimization module 280 is used to identify key process nodes or key connection edges whose carbon emission contribution exceeds a preset threshold based on the dynamic carbon footprint assessment results, generate supplier replacement, process parameter optimization, energy substitution and logistics switching scenarios for the key process nodes or key connection edges and perform multi-objective optimization, and output the preferred scenario scheme for updating the parameters of the life cycle process network.

[0271] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3As shown, this embodiment of the invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following steps:

[0272] Acquire activity data, energy consumption data, transportation data, and emission factor data of the target product during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use, and recycling and disposal, and perform field alignment, time synchronization, and unit conversion;

[0273] A lifecycle process network is constructed based on the processed data. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0274] Identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and determine the affected sub-networks along the life cycle process network starting from the node corresponding to the event;

[0275] Based on the event, perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network;

[0276] Artificial intelligence models are used to predict missing activity data, time-varying emission factors, and hidden process emissions in affected subnetworks;

[0277] Based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints, the prediction results are truncated, backtracked, or re-estimated, and the lifecycle inventory is updated.

[0278] The carbon emissions of the affected subnetworks are recalculated, and the historical accounting results of the unaffected parts are retrieved and summarized according to node identifier, stage identifier or timestamp to obtain the dynamic carbon footprint assessment results.

[0279] Based on the dynamic carbon footprint assessment results, key process nodes or key connections with carbon emission contributions exceeding a preset threshold are identified. For these key process nodes or key connections, scenarios such as supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated and multi-objective optimization is performed. The optimal scenario scheme for updating the parameters of the life cycle process network is then output.

[0280] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it performs the following steps:

[0281] Acquire activity data, energy consumption data, transportation data, and emission factor data of the target product during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use, and recycling and disposal, and perform field alignment, time synchronization, and unit conversion;

[0282] A lifecycle process network is constructed based on the processed data. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes.

[0283] Identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and determine the affected sub-networks along the life cycle process network starting from the node corresponding to the event;

[0284] Based on the event, perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network;

[0285] Artificial intelligence models are used to predict missing activity data, time-varying emission factors, and hidden process emissions in affected subnetworks;

[0286] Based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints, the prediction results are truncated, backtracked, or re-estimated, and the lifecycle inventory is updated.

[0287] The carbon emissions of the affected subnetworks are recalculated, and the historical accounting results of the unaffected parts are retrieved and summarized according to node identifier, stage identifier or timestamp to obtain the dynamic carbon footprint assessment results.

[0288] Based on the dynamic carbon footprint assessment results, key process nodes or key connections with carbon emission contributions exceeding a preset threshold are identified. For these key process nodes or key connections, scenarios such as supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated and multi-objective optimization is performed. The optimal scenario scheme for updating the parameters of the life cycle process network is then output.

[0289] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0290] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0291] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An AI-driven dynamic lifecycle assessment and scenario optimization method for product carbon footprint, characterized in that, include: Acquire activity data, energy consumption data, transportation data, and emission factor data of the target product during the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use, and recycling and disposal, and perform field alignment, time synchronization, and unit conversion; A lifecycle process network is constructed based on the processed data. The lifecycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes. Identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and determine the affected sub-networks along the life cycle process network starting from the node corresponding to the event; Based on the event, perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network; Artificial intelligence models are used to predict missing activity data, time-varying emission factors, and hidden process emissions in affected subnetworks; Based on the principles of mass conservation, energy conservation, process flow, and equipment capacity constraints, the prediction results are truncated, backtracked, or re-estimated, and the lifecycle inventory is updated. The carbon emissions of the affected subnetworks are recalculated, and the historical accounting results of the unaffected parts are retrieved and summarized according to node identifier, stage identifier or timestamp to obtain the dynamic carbon footprint assessment results. Based on the dynamic carbon footprint assessment results, key process nodes or key connections with carbon emission contributions exceeding a preset threshold are identified. For these key process nodes or key connections, scenarios such as supplier replacement, process parameter optimization, energy substitution, and logistics switching are generated and multi-objective optimization is performed. The optimal scenario scheme for updating the parameters of the life cycle process network is then output.

2. The method according to claim 1, characterized in that, The activity data, energy consumption data, transportation data, and emission factor data are respectively mapped to process nodes or connection edges in the life cycle process network; Specifically, raw material procurement data and supplier data are mapped to process nodes in the raw material acquisition stage; production equipment operation data and process energy consumption data are mapped to process nodes in the production and manufacturing stage; transportation operation data are mapped to transportation connection edges; energy consumption data in the usage stage is mapped to process nodes in the usage stage; recycling and disposal data are mapped to process nodes in the recycling and disposal stage; and emission factor data are mapped to corresponding process nodes.

3. The method according to claim 1, characterized in that, In the life cycle process network, each process node corresponds to a life cycle unit process, each connection edge corresponds to a material flow, energy flow or transportation flow, and each process node is associated with activity volume, emission factor, timestamp and stage identifier. The stage identifier is used to distinguish the raw material acquisition stage, production and manufacturing stage, warehousing and logistics stage, use stage and recycling and disposal stage.

4. The method according to claim 1, characterized in that, The affected sub-network is determined as follows: Starting from the node corresponding to the event, the propagation search is performed along the upstream and downstream connection edges in the life cycle process network until a preset influence depth or a preset influence threshold is reached, resulting in a set of affected process nodes and connection edges. The preset influence depth is the number of propagation levels starting from the node corresponding to the event, and the preset influence threshold is a threshold for the percentage change in node activity or a threshold for the percentage change in node carbon emissions.

5. The method according to claim 1, characterized in that, The inputs to the artificial intelligence model are historical activity data, real-time equipment parameters, process parameters, transportation distance, and historical emission factor sequences. The outputs of the artificial intelligence model are missing activity data, time-varying emission factors, and implicit process emissions. The implicit process emissions are process emissions that are not directly collected but inferred from equipment parameters or process status.

6. The method according to claim 1, characterized in that, The mass conservation, energy conservation, process flow, and equipment capacity constraints are used to verify whether the prediction results meet the material balance relationship, unit output energy consumption range, process sequence, and equipment rated processing capacity. Prediction results that exceed the equipment rated processing capacity or energy consumption range are truncated and corrected. Prediction results that do not meet the material balance relationship or energy balance relationship are backtracked and corrected. Prediction results that are missing key input variables are re-evaluated and corrected.

7. The method according to claim 1, characterized in that, The process of recalculating carbon emissions for the affected subnetworks involves retrieving historical calculation results from unaffected parts based on node identifiers, stage identifiers, or timestamps, and summarizing these results to obtain a dynamic carbon footprint assessment. Specifically: Extract the process nodes and connecting edges within the affected subnetwork and recalculate their carbon emissions. Read the historical carbon emissions corresponding to the unaffected parts according to the node identifier, stage identifier, or timestamp, and summarize the recalculation results and the historical carbon emissions.

8. The method according to claim 1, characterized in that, The identification of key process nodes or key connection edges whose carbon emission contribution exceeds a preset threshold specifically includes: Calculate the proportion of carbon emissions of each process node and connecting edge to the total carbon footprint of the product, and identify process nodes or connecting edges whose proportions are higher than a preset threshold as key process nodes or key connecting edges.

9. The method according to claim 1, characterized in that, The multi-objective optimization takes carbon emissions, production costs, delivery cycle and quality risks as optimization objectives. Under the constraints of supply capacity, equipment capacity and process feasibility, it optimizes the parameters corresponding to raw material supply nodes, process nodes, energy supply nodes or transportation connection edges, and determines the preferred scenario scheme for updating the parameters of the corresponding process nodes or connection edges in the life cycle process network.

10. An AI-driven dynamic lifecycle assessment and scenario optimization system for product carbon footprint, characterized in that, An AI-driven dynamic lifecycle assessment and scenario optimization method for the carbon footprint of a product as described in any one of claims 1 to 9, comprising: The data acquisition and preprocessing module is used to acquire activity data, energy consumption data, transportation data and emission factor data of the target product in the stages of raw material acquisition, production and manufacturing, warehousing and logistics, use and recycling and disposal, and to perform field alignment, time synchronization and unit conversion. The process network construction module is used to construct a life cycle process network based on the processed data. The life cycle process network includes process nodes and material flow connection edges, energy flow connection edges and transportation connection edges between nodes. The event identification module is used to identify events such as supplier changes, process parameter adjustments, energy structure changes, logistics route switching, and emission factor updates, and to determine the affected sub-network along the life cycle process network starting from the node corresponding to the event. The dynamic reconfiguration module is used to perform process node replacement, connection edge adjustment, system boundary adjustment, or emission factor update on the life cycle process network based on the event. The prediction module is used to predict missing activity data, time-varying emission factors, and implicit process emissions in the affected subnetworks using artificial intelligence models; The correction and update module is used to perform truncation correction, backtracking correction, or revaluation correction on the prediction results based on mass conservation, energy conservation, process flow, and equipment capacity constraints, and to update the lifecycle inventory. The incremental accounting module is used to recalculate carbon emissions for the affected sub-networks, retrieve historical accounting results of the unaffected parts according to node identifier, stage identifier or timestamp, and summarize them to obtain dynamic carbon footprint assessment results. The scenario optimization module is used to identify key process nodes or key connections with carbon emission contributions exceeding a preset threshold based on the dynamic carbon footprint assessment results. It generates supplier replacement, process parameter optimization, energy substitution, and logistics switching scenarios for the key process nodes or key connections and performs multi-objective optimization, outputting the preferred scenario scheme for updating the parameters of the life cycle process network.

Images