Green product evaluation model optimization method and system based on big data

Abstract

This invention provides a method and system for optimizing a green product evaluation model based on big data, relating to the field of big data technology. First, it collects reverse-traceability green attribute data throughout the entire lifecycle of green products, covering key data from multiple stages such as recycling, use, manufacturing, and raw material acquisition. Next, it reverse-engineers attribute relationships and traceability nodes between each stage, constructing a set of green attribute reverse calibration components that includes reverse traceability rules, node feature mapping relationships, and stage attribute feedback channels. Then, the data is input into the component set according to traceability nodes, and transmitted back to the previous stage evaluation stage through the stage attribute feedback channels, generating evaluation deviation trajectories and node feature mismatch information. Finally, based on this information, it dynamically adjusts the reverse traceability rules and the feature response thresholds of each stage evaluation stage of the evaluation model, obtaining an optimized green product evaluation model that improves the accuracy and reliability of the evaluation.

Description

Technical Field

[0001] This invention relates to the field of big data technology, and more specifically, to a method and system for optimizing a green product evaluation model based on big data. Background Technology

[0002] In today's society, with the continuous improvement of environmental awareness and the deepening of the concept of sustainable development, the research and development and promotion of green products have become the focus of attention for various industries. Green product assessment, as a crucial link in measuring the greenness of products and guiding their improvement and development, is of paramount importance in terms of accuracy and scientific rigor.

[0003] Currently, traditional green product assessment methods primarily focus on evaluating a specific stage or certain attributes of a product, such as energy consumption during the manufacturing stage or pollutant emissions during the usage stage. These assessment methods have significant limitations. Due to a lack of in-depth exploration of the complex relationships between different stages of the product's entire lifecycle, they fail to comprehensively and accurately reflect the true green attributes of the product. For example, when assessing products, the impact of the raw material acquisition stage on subsequent manufacturing and usage stages is not fully considered, as well as the inherent connection between the recycling and usage stages. This leads to discrepancies between the assessment results and the actual situation, failing to provide a reliable basis for the optimized design, production, and improvement of green products, thus hindering the healthy development of the green product industry. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for optimizing a green product evaluation model based on big data, the method comprising: Collect reverse traceability green attribute data for the entire life cycle of green products. The reverse traceability green attribute data includes data on the purity of recycled resources, energy consumption control data, actual energy consumption data, loss and degradation data, actual emission reduction data of the manufacturing process, raw material conversion efficiency data, environmental impact data of mining, and resource utilization rate data during the raw material acquisition stage. The reverse mining process is used to identify the attribute relationships and traceability nodes between the recycling stage and the usage stage, the usage stage and the production and manufacturing stage, and the production and manufacturing stage and the raw material acquisition stage in the reverse traceability green attribute data. The traceability nodes are the key feature intersection points in the reverse attribute transmission process. Based on the attribute association and tracing nodes, a set of green attribute reverse calibration components for the evaluation model is constructed. The set of green attribute reverse calibration components includes reverse tracing rules, node feature mapping relationships, and stage attribute feedback channels. The reverse traceability green attribute data is input into the green attribute reverse calibration component set according to the traceability node, and is transmitted back to the previous stage evaluation link of the evaluation model through the stage attribute feedback channel to generate green product evaluation deviation trajectory and node feature mismatch information. Based on the evaluation deviation trajectory and node feature mismatch information, the reverse tracing rules in the green attribute reverse calibration component set and the feature response thresholds of each stage of the evaluation model are dynamically adjusted to obtain the optimized green product evaluation model.

[0005] In another aspect, embodiments of the present invention also provide a green product evaluation model optimization system based on big data, including a processor and a machine-readable storage medium. The machine-readable storage medium is connected to the processor. The machine-readable storage medium is used to store programs, instructions, or code. The processor is used to run the programs, instructions, or code in the machine-readable storage medium to implement the above-described method.

[0006] Based on the above, this invention collects reverse-traceability green attribute data for the entire lifecycle of green products, covering detailed information from multiple key stages from raw material acquisition to recycling. This comprehensively and meticulously reflects the green characteristics of green products throughout their entire lifecycle, reverse-engineers attribute relationships and traceability nodes between stages, and accurately locates key feature intersections in the attribute transfer process, leading to a deeper understanding of the complex relationships between different stages of green products. Based on this, a set of green attribute reverse calibration components is constructed, including reverse traceability rules, node feature mapping relationships, and stage attribute feedback channels, enabling multi-dimensional calibration and optimization of the evaluation model. By inputting data according to traceability nodes and reverse-transmitting it to the preceding stage evaluation stage, a green product evaluation deviation trajectory and node feature mismatch information are generated, enabling timely identification of error sources and problems in the evaluation process. Finally, the reverse traceability rules and feature response thresholds of each stage of the evaluation model are dynamically adjusted to obtain an optimized green product evaluation model, effectively improving the accuracy and reliability of the evaluation. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the execution flow of the green product evaluation model optimization method based on big data provided in the embodiments of the present invention.

[0008] Figure 2 This is a schematic diagram of the hardware architecture of the green product evaluation model optimization system based on big data provided in this embodiment of the invention. Detailed Implementation

[0009] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1This is a flowchart illustrating a method for optimizing a green product evaluation model based on big data, as provided in one embodiment of the present invention. The following is a detailed description of this method.

[0010] Step S110: Collect reverse traceability green attribute data for the entire life cycle of green products. The reverse traceability green attribute data includes data on the purity of recycled resources, energy consumption control data, actual energy consumption data, loss and degradation data, actual emission reduction data of the manufacturing process, raw material conversion efficiency data, environmental impact data of mining in the raw material acquisition stage, and resource utilization rate data.

[0011] In this embodiment, lithium-ion batteries for electric vehicles are used as an example of green products to collect green attribute data through reverse traceability throughout the entire life cycle. During the recycling stage, purity data of recycled resources is collected through a distributed sensor network on the recycling processing line. This purity data is obtained by multi-dimensional component detection of the electrode materials of the recycled batteries, covering the purity index of key metal elements in the positive electrode active material, the structural integrity parameters of the negative electrode material, and the retention of physical properties of the separator. Recycling energy consumption control data is collected through metering devices such as smart meters and gas meters connected to each processing unit, including the unit material energy consumption of the crushing process, the energy conversion efficiency of the roasting process, and the energy consumption fluctuation coefficient in the hydrometallurgical process. Actual energy consumption data during the usage stage is collected through the real-time monitoring module of the battery management system (BMS), including the energy conversion efficiency at different charge / discharge rates, the self-discharge rate of the battery in different temperature ranges, and the battery energy output characteristics under vehicle operating conditions. Degradation data is obtained through the state of health (SOH) estimation model built into the BMS and periodic offline detection, including the cycle life decay curve of the electrode material, the trend of the ionic conductivity change of the electrolyte, and the growth characteristics of the battery's internal resistance. Actual emission reduction data during the manufacturing process are collected through the factory's environmental management system, covering carbon emission intensity during cathode material calcination, VOCs emission reduction from the organic solvent recovery system, and pollutant removal efficiency from the wastewater treatment unit. Raw material conversion efficiency data is collected through the material tracking module of the Manufacturing Execution System (MES), including the conversion rate from lithium source to cathode material, material balance coefficients for each process, and dynamic changes in material utilization during equipment operation. Environmental impact data during the raw material acquisition phase is collected through multi-parameter sensors at the mine's environmental monitoring station, including the ecological disturbance index of the mining area, soil physicochemical property changes, and water pollutant concentrations. Resource utilization data is collected through the mine resource management system, including mineral resource recovery rate, concentrate grade improvement rate during beneficiation, and resource conversion rate of tailings. During the data collection process, data anonymization technology is used for sensitive data such as mining area locations and specific production enterprise information. Sensitive fields are encrypted and replaced or obfuscated to ensure privacy protection during data transmission and storage. At the same time, end-to-end encryption protocols are used for data transmission to prevent data from being illegally intercepted and leaked during transmission.

[0012] Step S120: Reverse mining the attribute associations and traceability nodes in the reverse traceability green attribute data of the recycling stage and the usage stage, the usage stage and the production and manufacturing stage, and the production and manufacturing stage and the raw material acquisition stage. The traceability nodes are the key feature intersection points in the reverse attribute transmission process.

[0013] In this embodiment, based on the collected green attribute data of reverse traceability of electric vehicle lithium-ion batteries, attribute associations between different stages and reverse mining of traceability nodes are performed.

[0014] Step S121: Using the purity data of recycled resources and the energy consumption control data of the recycling stage as the starting data source, extract the attribute features related to the use stage. The extracted attribute features related to the use stage reflect the impact of the product status in the use stage on the product attributes in the recycling stage.

[0015] From the purity data of recycled resources during the recycling stage, feature extraction algorithms are used to identify attribute features related to the usage stage, such as the lattice structure integrity of metal elements in the cathode material. The data content of this lattice structure integrity feature reflects the impact of the structural fatigue of the electrode material during the charging and discharging process of the battery on the purity of the recycled material. From the energy consumption control data of recycling, abnormal energy consumption fluctuation features during the recycling process are extracted. The data content of this abnormal energy consumption fluctuation feature reflects that improper use of the battery during the usage stage, such as overcharging and discharging and high-temperature use, leads to damage to the internal structure of the battery, thereby affecting the energy consumption control effect during recycling.

[0016] Step S122: Track the transmission trajectory of the extracted attribute features related to the usage stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the actual energy consumption data and loss degradation data of the usage stage, and establish a preliminary attribute association between the recycling stage and the usage stage.

[0017] The data tracing algorithm is used to track the transmission trajectory of the attribute features related to the usage stage extracted in step S121.

[0018] Step S1221: Mark the extracted attribute features related to the usage stage. The marking content includes the attribute type of the attribute feature, the data source of the attribute feature, and the core expression form of the attribute feature, so that the attribute features are identifiable in the process of tracking the transmission trajectory.

[0019] The extracted "lattice structure integrity features of cathode material metal elements" are labeled as "structural integrity features", the data source is labeled as "recycling stage - cathode material detection", and the core manifestation is labeled as "lattice distortion rate - crystal interplanar spacing change - element distribution uniformity multidimensional feature vector". By adding specific feature identifiers to the data header, this attribute feature can be accurately identified in the subsequent trajectory tracking process.

[0020] Step S1222: Starting from the attribute data storage area of ​​the recycling stage in reverse chronological order, retrieve the attribute feature data stream with the attribute feature mark in the reverse tracing green attribute data, and record the storage location and associated data identifier of the attribute feature data stream.

[0021] Starting from the distributed database cluster in the attribute data storage area of ​​the recycling phase, following the reverse chronological order, and based on the feature identifier added in step S1221, a distributed retrieval engine performs a full scan of the reverse-source green attribute data to retrieve the attribute feature data stream marked with that attribute feature. The shard storage location, data block index information, and associated metadata identifiers, such as data collection timestamp, device number, and testing batch number, are recorded for this data stream in the distributed database.

[0022] Step S1223: Follow the retrieved attribute feature data stream, track the movement path of the attribute feature data stream in the reverse source green attribute data, record the intermediate storage units and data processing nodes through which the attribute feature data stream passes, and form a preliminary transmission trajectory.

[0023] Based on the storage location and data identifier recorded in step S1222, the movement path of the attribute feature data stream in the reverse-source green attribute data is followed using a data link tracing tool. The data stream is recorded from its origin in the data storage area of ​​the recycling stage, passing through intermediate data processing nodes such as data cleaning nodes, feature extraction nodes, and data fusion nodes, as well as intermediate storage units such as cache servers and message queues during transmission between nodes, forming a preliminary transmission trajectory map.

[0024] Step S1224: In the initial transmission trajectory, filter out the attribute feature data stream branches that point to the data storage area of ​​the usage stage, focus on the transmission details of the attribute feature data stream branches, and record the entry position of the attribute feature data stream into the data storage area of ​​the usage stage.

[0025] Branch analysis is performed on the preliminary transmission trajectory map formed in step S1223. Based on the data flow direction identifier, data flow branches pointing to the data storage area in the usage stage are selected. Focusing on this branch, detailed records are made of the last data processing node the data flow passes through before entering the data storage area in the usage stage, the data transmission protocol, and the specific entry point into the data storage area in the usage stage, such as the database table name and partition key value.

[0026] Step S1225: Within the data storage area during the usage phase, retrieve the attribute features in the actual energy consumption data according to the attribute feature markers, compare the attribute features in the retrieved actual energy consumption data with the attribute types and core manifestations of the marked attribute features, and find similar attribute feature items.

[0027] Within the data storage area during the usage phase, based on the attribute feature markings in step S1221, attribute features in the actual energy consumption data are retrieved using Structured Query Language (SQL) or unstructured data retrieval tools. The retrieved attribute features are compared with the attribute types and core manifestations of the marked attribute features, feature similarity is calculated, and attribute feature items with similarity higher than a set threshold are selected as similar attribute feature items.

[0028] Step S1226: Using the same method, retrieve the attribute features in the loss and degradation data within the data storage area during the usage phase, compare the attribute features in the retrieved loss and degradation data with the attribute types and core manifestations of the marked attribute features, and find similar attribute feature items.

[0029] Using the same retrieval and comparison method as in step S1225, attribute feature retrieval and similarity comparison are performed on the loss and degradation data in the data storage area during the usage phase to find attribute feature items similar to the marked attribute features.

[0030] Step S1227: Analyze the degree of correlation between similar attribute features and marked attribute features. By analyzing the continuity of the attribute feature transmission path and the correlation of attribute feature parameters, determine the attribute features in the actual energy consumption data and the attribute features in the loss and degradation data that directly correspond to the marked attribute features.

[0031] For the similar attribute features selected in steps S1225 and S1226, their correlation with the labeled attribute features is analyzed from two dimensions: continuity of transmission path and parameter correlation.

[0032] For example, step S1227-1: Track the transmission sub-trajectory of similar attribute feature items in the data storage area during the use stage, compare the connection between the transmission sub-trajectory and the main trajectory of the marked attribute feature from the recycling stage to the use stage, and present the continuous connection state of the two trajectories.

[0033] By using a data trajectory tracking tool, the transmission sub-trajectories of each similar attribute feature item within the data storage area during the usage phase are tracked separately. These sub-trajectories are then compared with the main trajectory of the marked attribute features from the recycling phase to the usage phase. The connection between the two trajectories in terms of data nodes and transmission links is checked to determine whether there are any discontinuities such as data breaks or path forks, thus presenting a continuous and interconnected state of the two trajectories.

[0034] Step S1227-2: If the trajectory is continuous and connected, analyze the consistency of the data format and attribute feature parameters at the trajectory connection points, and record the matching of attribute feature parameters at the connection points as the first basis for judging the degree of correlation.

[0035] When two trajectories are in a continuous and connected state, further analysis is conducted to determine whether the data formats at the trajectory connection point are compatible, such as whether the data encoding method, field length, and data type are consistent. At the same time, the attribute feature parameters at the connection point are compared, such as the dimension of the feature vector, the value range of key parameters, and the data change trend. The matching degree of the parameters is recorded and used as the primary basis for judging the degree of correlation.

[0036] Step S1227-3: Extract the core parameters of similar attribute feature items and the core parameters of labeled attribute features, compare the type of attribute feature parameters, the value range of attribute feature parameters, and the changing trend of attribute feature parameters, and deduce the degree of fit of attribute feature parameters as the second criterion for judging the degree of association.

[0037] The core parameters of similar attribute features and labeled attribute features are extracted separately. Matching analysis is performed on the parameter types to determine whether they are the same type of physical or chemical attribute parameters. The value ranges of the parameters are compared to analyze the degree of intersection between the two. Through trend analysis algorithms, the trend of parameter changes over time or other influencing factors is compared to see if they are consistent, and the degree of fit of the parameters is derived as a second criterion for judging the degree of correlation.

[0038] Step S1227-4: Analyze the correlation between similar attribute features and labeled attribute features in the time dimension, that is, the matching status between the generation time of similar attribute features and the transmission time of labeled attribute features, and record the time matching status as the third criterion for judging the degree of correlation.

[0039] Obtain the generation timestamps of similar attribute features and the transmission timestamps of labeled attribute features. Use time series analysis methods to analyze the temporal order and stability of the time interval between the two to determine whether there is a causal relationship. Record the time matching situation as a third criterion for judging the degree of association.

[0040] Step S1227-5: Find the associated data records of similar attribute feature items and labeled attribute features. If there is a clear attribute feature data flow pointing relationship, use the attribute feature data flow pointing relationship as the fourth criterion for judging the degree of association.

[0041] In the metadata of the reverse-source green attribute data, search for records indicating data flow relationships between similar attribute features and labeled attribute features, such as data lineage records and feature dependency graphs. If a clear relationship exists, use it as the fourth criterion for judging the degree of association.

[0042] Step S1227-6: Based on the four judgment criteria, comprehensively evaluate the degree of association between similar attribute features and labeled attribute features, and classify the association level.

[0043] Based on the four judgment criteria in steps S1227-2 to S1227-5, a weighted scoring method is used to comprehensively evaluate the degree of association between similar attribute features and labeled attribute features. The association level is divided according to the evaluation score, such as extremely high, high, medium, and low.

[0044] Step S1227-7: Filter out the similar attribute feature items with the highest correlation level and prioritize them as candidates for directly corresponding attribute feature items.

[0045] Based on the association level assessment results, the similar attribute features with the highest association level are selected and listed as candidate attribute features that directly correspond to the labeled attribute features.

[0046] Step S1227-8: Verify the integrity of the transmission path of the candidate attribute feature items, and show that the candidate attribute feature items can be traced back to the original data source of the labeled attribute feature through continuous trajectory.

[0047] The integrity of the transmission path of candidate attribute features is verified. Starting from the data storage area of ​​the candidate attribute feature in the usage phase, the transmission path is traced backward to check whether it is possible to trace back to the original data source of the marked attribute feature in the recycling phase through a continuous trajectory path, ensuring that the transmission path is uninterrupted and without omissions.

[0048] Step S1227-9: Confirm that the function of the candidate attribute feature in the evaluation during the use phase has an inverse correlation with the function of the marked attribute feature in the recycling phase. The inverse correlation is manifested in that: the increase or decrease of the value of the candidate attribute feature will lead to the expected change of the evaluation value of the marked attribute feature in the opposite or same direction.

[0049] This study analyzes the functional roles of candidate attribute features in the usage phase evaluation and the functional roles of labeled attribute features in the retrieval phase. Using a functional correlation analysis model, it determines whether there is an inverse correlation between the two. Specifically, when the value of a candidate attribute feature changes, does the evaluation value of the labeled attribute feature exhibit the expected positive or negative change? For example, does an increase in the value of a candidate attribute feature lead to an increase in the evaluation value of the labeled attribute feature, or vice versa?

[0050] Step S1227-10: When a candidate attribute feature item satisfies the condition that the transmission path is complete and there is the reverse correlation with the marked attribute feature item, it is finally determined as the attribute feature item in the actual energy consumption data and the attribute feature item in the loss degradation data that directly correspond to the marked attribute feature item.

[0051] If a candidate attribute feature passes the transmission path integrity verification and has the aforementioned reverse correlation with the marked attribute feature, it will be ultimately determined as an attribute feature in the actual energy consumption data or loss degradation data of the usage stage that directly corresponds to the marked attribute feature.

[0052] Step S1228: Record the storage address, associated data group, and data generation time of the corresponding attribute feature item in the data storage area during the usage phase to form attribute feature item location information.

[0053] For each identified attribute feature, record its specific storage address within the data storage area during the usage phase, such as the database's IP address, port number, database name, table name, row key, etc.; associated data group information, such as the identifiers of other related attribute features, data collection batch numbers, etc.; and the timestamp of data generation, forming complete attribute feature location information.

[0054] Step S1229: Associate and integrate the attribute feature markers, attribute feature transmission trajectories, and attribute feature item location information, and extract and present the complete correspondence between the attribute features extracted in the recycling stage and those related to the usage stage, the attribute features in the actual energy consumption data of the usage stage, and the attribute features in the loss and degradation data.

[0055] By associating and integrating the attribute feature markings in step S1221, the attribute feature transfer trajectory formed in step S1223, and the attribute feature item location information formed in step S1228, a correspondence map of attribute features between the recycling stage and the usage stage is constructed, clearly presenting the complete correspondence between the attribute features extracted in the recycling stage and the corresponding attribute feature items in the actual energy consumption data and loss degradation data in the usage stage.

[0056] Step S12210: Based on the complete correspondence, supplement and improve the details of the attribute feature transmission trajectory, and mark the data form changes of the attribute feature items during the transmission process.

[0057] Based on the complete correspondence obtained through integration, the attribute feature transmission trajectory is supplemented and improved in detail, and the data form changes of attribute feature items when passing through each data processing node during the transmission process are clearly marked, such as data format conversion, feature dimension increase or decrease, parameter value transformation, etc.

[0058] Step S123: Analyze the interaction of attribute features in the preliminary attribute association, identify the core attribute features that remain stable during the transmission of attribute features, and determine the intersection of the core attribute features as the first traceability node between the recycling stage and the usage stage.

[0059] A thorough analysis is conducted on the preliminary attribute associations established in step S122 between the recycling and usage stages. Using a feature interaction analysis model, the interaction mechanisms between various attribute features are analyzed, such as promoting, inhibiting, and synergistic effects. Based on this, core attribute features whose characteristic parameters and manifestations remain relatively stable during the reverse transmission of attribute features from the usage stage to the recycling stage are identified. Through data node positioning technology, the intersection points of these core attribute features in the reverse-source green attribute data are determined; these intersection points constitute the first traceability node between the recycling and usage stages.

[0060] Step S124: Using the actual energy consumption data and loss degradation data of the usage stage as intermediate data sources, extract the attribute features related to the production and manufacturing stage. The extracted attribute features related to the production and manufacturing stage reflect the impact of the process parameters of the production and manufacturing stage on the product attributes of the usage stage.

[0061] From the actual energy consumption data during the usage phase, attribute features related to the manufacturing phase are extracted using a feature selection algorithm. For example, the battery's charge and discharge efficiency features reflect the impact of process parameters such as the coating uniformity and compaction density of electrode materials during the manufacturing phase on the battery's energy conversion efficiency during the usage phase. From the loss and degradation data, the battery's capacity decay rate features are extracted. The data reflects the impact of process parameters such as the purity of the electrolyte and the additive ratio during the manufacturing phase on the battery's performance degradation during the usage phase.

[0062] Step S125: Track the transmission trajectory of the extracted attribute features related to the production and manufacturing stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the actual emission reduction data and raw material conversion efficiency data of the production and manufacturing stage, and establish the attribute association between the use stage and the production and manufacturing stage.

[0063] Using a trajectory tracing method similar to step S122, the attribute features related to the production and manufacturing stage extracted in step S124 are traced in the reverse traceability green attribute data. Starting from the data storage area of ​​the usage stage, the data is traced in reverse chronological order to locate the attribute feature item corresponding to the attribute feature in the actual emission reduction data and raw material conversion efficiency data of the production and manufacturing stage, thereby establishing the attribute association between the usage stage and the production and manufacturing stage.

[0064] Step S126: Analyze the transmission logic of attribute feature items in the attribute association, identify the key transformation attribute features when attribute feature items are transferred from the manufacturing stage to the usage stage, and determine the intersection of the key transformation attribute features as the second traceability node between the usage stage and the manufacturing stage.

[0065] The transmission logic of attribute feature items in the attribute association between the usage stage and the manufacturing stage is analyzed.

[0066] Step S1261: Decompose all attribute feature items in the attribute association between the usage stage and the manufacturing stage, classify them according to the transmission direction of the attribute feature items, and distinguish between attribute feature items output from the manufacturing stage and attribute feature items input to the usage stage.

[0067] All attribute features associated with the usage stage and the manufacturing stage are broken down and divided into two main categories based on the direction of data transmission: attribute features output from the manufacturing stage and attribute features input to the usage stage.

[0068] Step S1262: Compare the attribute parameters of the attribute features output in the manufacturing stage with those of the attribute features input in the usage stage, and analyze the internal logic of the changes in attribute parameters. The internal logic describes the transformation process of attribute features from the manufacturing stage to the usage stage attribute features based on predefined process transformation rules and product form mapping relationships.

[0069] By comparing the attribute parameters of the output attribute features in the manufacturing stage with those of the input attribute features in the usage stage, and combining predefined lithium-ion battery production process transformation rules (such as the influence rules of coating, rolling, assembly and other processes on battery performance) and product form mapping relationships (such as the mapping relationship between electrode sheet form and battery cell performance), the inherent logic of the changes in attribute parameters from the manufacturing stage to the usage stage is analyzed, and it is clarified how the attribute features in the manufacturing stage are transformed into the attribute features in the usage stage through process transformation and form mapping.

[0070] Step S1263: Track the transformation process of output attribute feature items in each production and manufacturing stage, record the key operations of attribute parameter adjustment, attribute feature addition and attribute feature merging in the transformation process, and construct the transformation process record.

[0071] For each production and manufacturing stage, the attribute feature items output are traced from the data storage area of ​​the production and manufacturing stage to the attribute feature items of the usage stage. The key operations experienced during the transformation process are recorded, such as attribute parameter adjustment (e.g., numerical scaling, unit conversion), attribute feature addition (e.g., new features generated due to process transformation), and attribute feature merging (e.g., merging multiple production features into one usage feature), to build a detailed transformation process record.

[0072] Step S1264: In the conversion process record, identify the conversion operation that plays a decisive role in the formation of the attribute characteristics of the usage stage. The attribute characteristic change corresponding to the conversion operation is the key conversion attribute characteristic.

[0073] Key operations are identified in the conversion process records. Sensitivity analysis is used to determine the conversion operations that play a decisive role in the formation of the attribute characteristics at the usage stage; that is, even a small change in this operation will lead to a significant change in the attribute characteristics at the usage stage. The attribute characteristic changes caused by this key conversion operation are the key conversion attribute characteristics.

[0074] Step S1265: Determine the location of key conversion attribute features in the conversion process record, and record the attribute feature data status, conversion operation type, and associated attribute feature items corresponding to that location.

[0075] In the conversion process record, accurately locate the first occurrence of the key conversion attribute feature, record the data status of the attribute feature at that location, such as data format, feature value range, etc.; the type of conversion operation, such as numerical calculation, logical judgment, feature extraction, etc.; and other attribute features related to the key conversion attribute feature.

[0076] Step S1266: Analyze the relationship between key conversion attribute features and other conversion attribute features, and extract and present how key conversion attribute features affect the conversion direction and degree of other attribute features.

[0077] By using association rule mining algorithms, we analyze the relationship between key conversion attribute features and other conversion attribute features in the conversion process records, and clarify how changes in key conversion attribute features affect the conversion direction (such as promoting or inhibiting) and degree of conversion (such as conversion rate and conversion magnitude) of other attribute features.

[0078] Step S1267: Locate the storage location of key transformation attribute features in the reverse traceability green attribute data, and search for related attribute feature data around that location. The related attribute feature data includes related production and manufacturing stage attribute feature data and usage stage attribute feature data.

[0079] By using data indexing technology, the physical storage location of key transformation attribute features in the reverse traceability green attribute data is located. Within the preset data range at this location, the associated production and manufacturing stage attribute feature data and usage stage attribute feature data are searched. These data together constitute the contextual information of the key transformation attribute features.

[0080] Step S1268: Determine the intersection area of ​​the associated attribute feature data, which contains the complete transformation trajectory of the key transformation attribute features and the interaction information of the associated attribute feature items.

[0081] Spatial distribution analysis is performed on the identified related attribute feature data to determine the intersection area of ​​these data in the reverse traceability green attribute data. This intersection area should include the complete transformation trajectory of key transformation attribute features from the manufacturing stage to the usage stage, as well as interaction information with other related attribute feature items, such as data exchange frequency and interaction intensity.

[0082] Step S1269: Define the intersection area as the second traceability node between the usage stage and the manufacturing stage, and extract and present the interactive data of the coverage of the second traceability node, which includes key transformation attribute features and related attribute features.

[0083] The intersection area determined in step S1268 is defined as the second traceability node between the usage stage and the manufacturing stage, and its coverage is defined. This coverage should be able to fully include the interaction data of key transformation attribute features and all related attribute feature items, so as to ensure that relevant feature information can be fully captured during subsequent reverse calibration.

[0084] Step S12610: Record the node identifier, coverage, core key transformation attribute features, and related attribute features of the second traceability node.

[0085] Assign a unique node identifier to the second traceability node, record its coverage in the reverse traceability green attribute data (such as the start and end addresses of data storage, the data tables involved, etc.), clarify the identifiers of the core key transformation attribute features, and a list of all related attribute feature items.

[0086] Step S127: Using the actual emission reduction data and raw material conversion efficiency data of the production and manufacturing stage as intermediate data sources, extract the attribute features related to the raw material acquisition stage. The extracted attribute features related to the raw material acquisition stage reflect the impact of resource data of the raw material acquisition stage on product attributes of the production and manufacturing stage.

[0087] From the actual emission reduction data of the production and manufacturing process, we extract attribute characteristics related to the raw material acquisition stage, such as the carbon emission coefficient characteristics of the cathode material production process. The data content reflects the impact of resource data such as the mining method and transportation distance of lithium ore in the raw material acquisition stage on the carbon emission level of the production and manufacturing stage. From the raw material conversion efficiency data, we extract the purity characteristics of the raw materials. The data content reflects the impact of resource data such as mineral processing and purification technology in the raw material acquisition stage on the raw material utilization rate in the production and manufacturing stage.

[0088] Step S128: Track the transmission trajectory of the extracted attribute features related to the raw material acquisition stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the mining environmental impact data and resource utilization rate data of the raw material acquisition stage, and establish the attribute association between the production and manufacturing stage and the raw material acquisition stage.

[0089] Using a trajectory tracking method similar to step S122, the extracted attribute features related to the raw material acquisition stage are traced back to the mining environment impact data and resource utilization data of the raw material acquisition stage, starting from the data storage area of ​​the production and manufacturing stage, and tracing back to the raw material acquisition stage to locate the corresponding attribute feature items, thereby establishing the attribute association between the production and manufacturing stage and the raw material acquisition stage.

[0090] Step S129: Analyze the interaction of attribute features in the attribute association, identify the core attribute features that remain stable during the transmission of attribute features, and determine the intersection of the core attribute features as the third traceability node between the production and manufacturing stage and the raw material acquisition stage.

[0091] The attribute correlation between the manufacturing stage and the raw material acquisition stage is analyzed. Through feature stability analysis, core attribute features whose characteristics and parameter values ​​remain relatively stable during the reverse transmission of attribute features from the raw material acquisition stage to the manufacturing stage are identified. The intersection of these core attribute features in the reverse traceability green attribute data is determined, and this intersection is the third traceability node between the manufacturing stage and the raw material acquisition stage.

[0092] Step S1210: Integrate the attribute associations between the recycling stage and the usage stage, the attribute associations between the usage stage and the manufacturing stage, the attribute associations between the manufacturing stage and the raw material acquisition stage, as well as the first traceability node, the second traceability node, and the third traceability node, to form a reverse attribute association and traceability node set.

[0093] The attribute relationships between the three stages mentioned above, as well as the information of the first, second, and third source nodes, are systematically integrated to construct a reverse attribute relationship graph, which includes information such as the relationship, relationship strength, and transmission direction between attribute feature items of each stage; at the same time, a set of source nodes is formed, recording detailed information such as the identifier, location, core attribute features, and related attribute feature items of each source node.

[0094] Step S130: Based on the attribute association and traceability nodes, construct a set of green attribute reverse calibration components for the evaluation model. The set of green attribute reverse calibration components includes reverse traceability rules, node feature mapping relationships, and stage attribute feedback channels.

[0095] In this embodiment, based on the reverse attribute association and traceability node set formed in step S1210, a set of green attribute reverse calibration components for the evaluation model of electric vehicle lithium-ion batteries is constructed.

[0096] Step S131: Analyze the attribute features of the first, second, and third source nodes, and extract and present the core attribute feature type, core attribute feature transmission direction, and core attribute feature interaction method of each source node.

[0097] The attribute features of the first tracing node are analyzed, and its core attribute features are extracted using a feature extraction algorithm. The types of core attribute features are determined, such as structural features and performance features. The transmission direction of the core attribute features is clarified, i.e., from which stage to which stage in reverse transmission. The interaction methods between core attribute features are analyzed, such as data sharing, parameter coupling, and feedback adjustment, and this information is presented in a structured form. The same analysis process is performed on the second and third tracing nodes.

[0098] Step S132: To address the reverse propagation requirement from the recycling phase to the usage phase, generate a first reverse tracing rule. The first reverse tracing rule includes the method of reverse mapping of attribute features from the recycling phase to the usage phase, the attribute feature matching criteria, and the attribute feature propagation priority.

[0099] Based on the reverse transmission requirements from the recycling phase to the usage phase, and combined with the attribute feature analysis results of the first traceability node, a first reverse traceability rule is generated. The reverse mapping method is defined as how the attribute features of the recycling phase are mapped to the attribute feature reference values ​​of the usage phase through mathematical transformations, logical reasoning, etc., such as weighted mapping based on feature similarity, regression mapping based on historical data, etc. The attribute feature matching standard specifies the conditions that must be met when matching attribute features of the recycling phase with those of the usage phase, such as quantitative indicators like feature type consistency, data range overlap, and similarity of change trends. The attribute feature transmission priority clarifies the order in which multiple attribute features from the recycling phase are transmitted to the usage phase, and is usually determined based on factors such as the degree of influence of the feature on the evaluation results and the stability of the feature.

[0100] Step S133: For the reverse transmission requirement from the usage stage to the production and manufacturing stage, generate a second reverse traceability rule. The content of the second reverse traceability rule includes the method of reverse mapping of the attribute features of the usage stage to the production and manufacturing stage, the attribute feature matching standard, and the attribute feature transmission priority.

[0101] Similar to step S132, for the reverse transmission requirement from the usage stage to the manufacturing stage, a second reverse traceability rule is generated based on the parsing results of the second traceability node. The reverse mapping method can employ modified mapping based on key transformation attribute features, mapping based on process parameter sensitivity, etc.; the attribute feature matching criteria consider the process correlation and parameter correspondence of attribute features between the usage stage and the manufacturing stage; the transmission priority is determined based on the importance of the usage stage attribute features to the evaluation of the manufacturing process.

[0102] Step S134: To address the reverse transmission requirement from the manufacturing stage to the raw material acquisition stage, a third reverse traceability rule is generated. The content of the third reverse traceability rule includes the method of reverse mapping of the attribute features of the manufacturing stage to the raw material acquisition stage, the attribute feature matching standard, and the attribute feature transmission priority.

[0103] To address the reverse transmission requirement from the manufacturing stage to the raw material acquisition stage, a third reverse traceability rule is generated based on the analysis results of the third traceability node. The reverse mapping method can employ proportional mapping based on raw material quality benchmarks, or correlation mapping based on resource utilization efficiency, etc.; the attribute feature matching standard focuses on the resource correlation and environmental impact correlation of attribute features between the manufacturing stage and the raw material acquisition stage; the transmission priority is determined based on the criticality of the manufacturing stage's attribute features to the assessment of the raw material acquisition stage.

[0104] Step S135: Establish node attribute feature mapping relationship, extract and present the correspondence between the core attribute features of the first traceability node and the attribute features of the usage stage evaluation link, the correspondence between the core attribute features of the second traceability node and the attribute features of the production and manufacturing stage evaluation link, and the correspondence between the core attribute features of the third traceability node and the attribute features of the raw material acquisition stage evaluation link.

[0105] In this embodiment, the node attribute feature mapping relationship is established through the following steps.

[0106] Step S1351: Extract the core attribute features of the first traceability node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the first traceability node.

[0107] From the attribute feature data of the first traceability node, core attribute features are extracted using a core feature extraction algorithm. For each core attribute feature, its attribute parameters are extracted in detail, such as parameter name, physical meaning, and dimensions; its representation format is clarified, such as numerical, vector, and matrix types; its data type is determined, such as integer, floating-point, and string types; and the above information is compiled into a list of core attribute features of the first traceability node.

[0108] Step S1352: Obtain all evaluation attribute features in the usage phase evaluation process. The evaluation objectives, data requirements, and parameter ranges of each evaluation attribute feature are extracted and presented from the design documents and data records of the corresponding evaluation phase, forming a list of evaluation attribute features for the usage phase.

[0109] By reviewing the design documents, data dictionary, and evaluation index system of the stage evaluation phase in the electric vehicle lithium-ion battery evaluation model, all evaluation attribute characteristics of this phase were obtained. For each evaluation attribute characteristic, its evaluation objective was clarified, i.e., which aspect of performance is this characteristic used to evaluate during the usage phase; data requirements, i.e., the raw data type, data precision, and data collection frequency required to evaluate this characteristic; parameter range, i.e., the normal value range and warning range of this characteristic, etc., and a list of evaluation attribute characteristics for the usage phase was compiled.

[0110] Step S1353: Compare the core attribute feature list of the first traceability node with the evaluation attribute feature list of the usage stage, find attribute feature combinations with consistent attribute parameter types and matching data types, and form preliminary corresponding candidate pairs.

[0111] A feature comparison algorithm is used to compare the attribute features in the core attribute feature list of the first traceability node and the attribute feature list of the usage stage evaluation item by item. The comparison focuses on whether the attribute parameter types are consistent (e.g., both are similar physical quantities such as length, mass, and concentration) and whether the data types match (e.g., both are floating-point or integer types). Attribute features that meet these conditions are combined to form preliminary candidate pairs.

[0112] Step S1354: Analyze the functional correlation between the core attribute features of the first traceability node and the evaluation attribute features of the usage stage in the preliminary corresponding candidate pairs. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the first traceability node and the data change trend of the evaluation attribute features of the usage stage. Valid corresponding candidate pairs are selected based on this correlation.

[0113] Functional correlation analysis is performed on the attribute features of the preliminary candidate pairs. By calculating the correlation coefficient of the data change trends of the two pairs, it is determined whether there is a positive correlation (if one feature increases, the other also increases) or a negative correlation (if one feature increases, the other decreases). Based on the preset correlation coefficient threshold, candidate pairs with significant functional correlation are selected as valid candidate pairs.

[0114] Step S1355: Determine the mapping method of valid corresponding candidate pairs, extract and present how the core attribute features of the first traceability node are transformed into calibration reference data for the evaluation attribute features in the usage stage, and form the first set of attribute feature correspondences.

[0115] For valid candidate pairs, based on the reverse mapping method in the first reverse tracing rule, a specific mapping function or mapping rule is determined, clarifying how the core attribute features of the first tracing node are transformed into calibration reference data for evaluating attribute features during the usage phase through this mapping method. For example, if the mapping method is linear, the slope, intercept, and other parameters of the mapping are determined; if it is nonlinear, the function type and parameters of the mapping are determined. The above mapping relationships are presented in the form of formulas, rule descriptions, etc., forming the first set of attribute feature correspondences.

[0116] Step S1356: Extract the core attribute features of the second traceability node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the second traceability node.

[0117] Using the same method as in step S1351, extract the core attribute features of the second traceability node, clarify its attribute parameters, expression form, and data type, and form a list of core attribute features of the second traceability node.

[0118] Step S1357: Obtain all evaluation attribute features of the evaluation stage in the production and manufacturing phase. The evaluation objectives, data requirements, and parameter ranges of each evaluation attribute feature are extracted and presented through the design documents and data records of the corresponding evaluation stage, forming a list of evaluation attribute features for the production and manufacturing phase.

[0119] Referring to step S1352, obtain all assessment attribute characteristics of the production and manufacturing stage assessment process, clarify its assessment objectives, data requirements, and parameter ranges, and form a list of assessment attribute characteristics for the production and manufacturing stage.

[0120] Step S1358: Compare the core attribute feature list of the second traceability node with the evaluation attribute feature list of the manufacturing stage, find attribute feature combinations with consistent attribute parameter types and matching data types, and form preliminary corresponding candidate pairs.

[0121] Using a feature comparison method similar to step S1353, the core attribute feature list of the second traceability node is compared with the evaluation attribute feature list of the production and manufacturing stage to form preliminary corresponding candidate pairs.

[0122] Step S1359: Analyze the functional correlation between the core attribute features of the second traceability node and the evaluation attribute features of the production and manufacturing stage in the preliminary corresponding candidate pairs. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the second traceability node and the data change trend of the evaluation attribute features of the production and manufacturing stage. Based on this correlation, select effective corresponding candidate pairs, determine the mapping method, and form the second set of attribute feature correspondences.

[0123] Functional correlation analysis is performed on the initial candidate pairs to select valid candidate pairs, and the mapping method is determined according to the second reverse tracing rule to form the second set of attribute feature correspondences.

[0124] Step S13510: Extract the core attribute features of the third source node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the third source node.

[0125] In the same step S1351, extract the core attribute features of the third source node and form a list.

[0126] Step S13511: Obtain all evaluation attribute features of the raw material acquisition stage evaluation process. The evaluation objectives, data requirements, and parameter ranges of each evaluation attribute feature are extracted and presented through the design documents and data records of the corresponding evaluation process, forming a list of evaluation attribute features for the raw material acquisition stage.

[0127] Referring to step S1352, obtain the evaluation attribute characteristics of the raw material acquisition stage and form a list.

[0128] Step S13512: Compare the core attribute feature list of the third traceability node with the evaluation attribute feature list of the raw material acquisition stage, identify attribute feature combinations with consistent attribute parameter types and matching data types, and form preliminary corresponding candidate pairs.

[0129] Preliminary candidate pairs are formed using feature comparison methods.

[0130] Step S13513: Analyze the functional correlation between the core attribute features of the third traceability node and the evaluation attribute features of the raw material acquisition stage in the preliminary corresponding candidate pairs. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the third traceability node and the data change trend of the evaluation attribute features of the raw material acquisition stage. Based on this correlation, select effective corresponding candidate pairs, determine the mapping method, and form the third set of attribute feature correspondence.

[0131] Analyze the functional correlation, screen valid candidate pairs, determine the mapping method, and form a third set of attribute feature correspondences.

[0132] Step S13514: Integrate the first set of attribute feature correspondences, the second set of attribute feature correspondences, and the third set of attribute feature correspondences to form a node feature mapping relationship.

[0133] The three sets of attribute feature correspondences formed in steps S1355, S1359, and S13513 are integrated to construct a unified node feature mapping table or mapping relationship map, which clearly shows the mapping relationship between the core attribute features of each traceability node and the corresponding stage evaluation attribute features.

[0134] Step S136: Plan and construct the first-stage attribute feedback channel, which connects the attribute data storage area of ​​the recycling stage with the evaluation stage of the usage stage. The first-stage attribute feedback channel is equipped with attribute feature transmission buffer processing, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the first reverse tracing rule.

[0135] The physical and logical architectures of the first-phase attribute feedback channel are planned. The physical architecture includes data transmission hardware and network topology, ensuring a smooth and efficient data transmission path from the attribute data storage area in the recycling phase to the evaluation phase in the usage phase. The logical architecture includes data transmission protocols, data encapsulation formats, and error handling mechanisms. An attribute feature transmission buffer processing unit is set up within the channel. This unit temporarily stores and controls the flow of transmitted attribute feature data to handle sudden fluctuations in data transmission. The data format (e.g., byte order, field arrangement) and transmission rate (e.g., number of data packets transmitted per unit time, data size) of the buffer processing unit must be strictly configured according to the data processing requirements defined in the first reverse tracing rule to ensure compatibility with upstream and downstream data processing nodes.

[0136] Step S137: Plan and construct the second-stage attribute feedback channel, which connects the usage stage attribute data storage area with the production and manufacturing stage evaluation link. The second-stage attribute feedback channel is equipped with attribute feature transmission buffer processing, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the second reverse tracing rule.

[0137] Referring to the planning and construction method in step S136, plan and construct the second-stage attribute feedback channel connecting the usage stage attribute data storage area with the production and manufacturing stage evaluation stage. The buffer processing unit configuration within the channel must meet the data processing requirements defined by the second reverse tracing rule, including parameters such as data format and transmission rate.

[0138] Step S138: Plan and construct the third-stage attribute feedback channel, which connects the attribute data storage area of ​​the production and manufacturing stage with the evaluation link of the raw material acquisition stage. The attribute feature transmission buffer processing is set in the third-stage attribute feedback channel, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the third reverse traceability rule.

[0139] Similarly, a third-stage attribute feedback channel is planned and constructed to connect the attribute data storage area of ​​the production and manufacturing stage with the evaluation link of the raw material acquisition stage, and the buffer processing unit is configured to meet the data processing requirements of the third reverse traceability rule.

[0140] Step S139: Establish a reverse calibration attribute feature integration process. This reverse calibration attribute feature integration process summarizes the reverse attribute feature data transmitted by the attribute feedback channels in the three stages, and sorts and integrates the attribute features according to the order of the traceability nodes.

[0141] A reverse calibration attribute feature integration process is designed. This process receives reverse attribute feature data from the output of three-stage attribute feedback channels. First, the data undergoes format validation and integrity checks to ensure data quality. Then, the received attribute feature data is sorted according to the order of the first traceability node, the second traceability node, and the third traceability node. Next, based on the node feature mapping relationship, attribute feature data transmitted from different channels that belong to the same traceability node or have a correlation are integrated, such as merging duplicate data, supplementing missing data, and correcting conflicting data, to form a unified reverse calibration attribute feature dataset.

[0142] Step S1310: Systematically integrate the first reverse tracing rule, the second reverse tracing rule, the third reverse tracing rule, the node feature mapping relationship, the first stage attribute feedback channel, the second stage attribute feedback channel, the third stage attribute feedback channel, and the reverse calibration attribute feature integration process to form a green attribute reverse calibration component set.

[0143] By employing component-based integration technology, the generated reverse tracing rules, node feature mapping relationships, constructed stage attribute feedback channels, and established reverse calibration attribute feature integration processes are systematically integrated. Interface specifications, data interaction protocols, and control logic between each component are defined to ensure collaborative operation. Ultimately, this results in an independent, reusable set of green attribute reverse calibration components, which, as part of the evaluation model, is responsible for implementing the reverse calibration function for attribute data.

[0144] Step S140: Input the reverse traceability green attribute data into the green attribute reverse calibration component set according to the traceability node, and transmit it back to the previous stage evaluation link of the evaluation model through the stage attribute feedback channel to generate green product evaluation deviation trajectory and node feature mismatch information.

[0145] The collected green attribute data of lithium-ion batteries for electric vehicles is input into the green attribute reverse calibration component set according to the classification of the first, second, and third traceability nodes. The attribute data of each traceability node is transmitted to the preceding stage evaluation stage of the evaluation model through the corresponding stage attribute feedback channel, following the reverse path of recycling stage → usage stage → manufacturing stage → raw material acquisition stage. In the preceding stage evaluation stage, the transmitted reverse attribute feature data is compared and analyzed with the evaluation attribute feature data currently used in that stage. By calculating the deviation value between the two (such as absolute deviation, relative deviation, root mean square deviation, etc.), and recording the change of the deviation value over time or evaluation cycle, a green product evaluation deviation trajectory is formed. Simultaneously, the causes of the deviation are analyzed to identify node feature mismatch information, such as attribute feature type mismatch, feature parameter range exceeding expectations, and abnormal feature interaction relationships.

[0146] Step S150: Based on the evaluation deviation trajectory and node feature mismatch information, dynamically adjust the reverse tracing rules in the green attribute reverse calibration component set and the feature response thresholds of each stage of the evaluation model to obtain the optimized green product evaluation model.

[0147] In this embodiment, based on the evaluation deviation trajectory and node feature mismatch information generated in step S140, dynamic adjustments are made by constructing a deviation-rule mapping matrix and a mismatch-threshold association model. The deviation-rule mapping matrix associates the deviation type and deviation value in the evaluation deviation trajectory with the reverse tracing rule clauses. The matrix element values ​​represent the correlation strength between the deviation and the rule clause, and the priority of the rule clauses to be adjusted is determined by ranking the correlation strength. The mismatch-threshold association model uses a multiple linear regression method, with the mismatch degree and mismatch feature dimension in the node feature mismatch information as independent variables and the feature response threshold as the dependent variable, to establish a quantitative correlation equation for calculating the threshold adjustment amount.

[0148] Step S151: Analyze and evaluate the deviation trajectory, locate the distribution of deviation at the first, second, and third traceability nodes, and extract and present the deviation type and deviation transmission direction at each traceability node.

[0149] Start the deviation analysis engine and load the evaluation deviation trajectory data and the reverse attribute association graph. Segment the evaluation deviation trajectory using the sliding window algorithm. The window size is set to K times the traceability node feature interaction period (K is an empirical coefficient calibrated according to historical data), and the step size is 1 / M of the window size (M is a positive integer). Perform frequency domain transformation on the deviation data of each window, extract the main frequency deviation features, and compare them with the preset main frequency library of node features to obtain the distribution ratio of deviations at the three traceability nodes. Calculate the deviation transfer coefficient for each traceability node. When the coefficient is positive and its absolute value is greater than N (N is a set threshold, 0 < N < 1), the deviation transfer direction is forward transfer; when the coefficient is negative and its absolute value is greater than N, it is reverse transfer. Determine the deviation type through the kurtosis value of the deviation probability density function: kurtosis value > P is impulse-type deviation, -P < kurtosis value ≤ P is normal-type deviation, and kurtosis value ≤ -P is flat-type deviation (P is the kurtosis value of the standard normal distribution). Present the results in a deviation distribution heat map and a transfer direction digraph.

[0150] Step S152: For the deviation at the first traceability node, analyze the degree of契合 between the attribute feature matching criteria in the first reverse traceability rule and the actual attribute features, and find the rule clause that causes the deviation.

[0151] Call the rule matching degree calculation module and input the first reverse traceability rule text and the actual attribute feature data stream of the first traceability node. Parse the rule text into a structured matching condition tree, and the leaf nodes are specific matching criteria (such as feature type matching degree ≥ A%, A is a rule-set value). Extract parameters such as the feature type vector, data form matrix, and change trend coefficient set from the actual attribute feature data stream. Through the condition tree traversal algorithm, compare the actual parameters with the rule conditions layer by layer and calculate the matching degree score for each layer (with a full score of Q points, Q is a positive integer). When the score of a certain layer < R points (R is a qualified threshold, R < Q), trigger the clause positioning mechanism, trace back the corresponding rule clause ID, and determine the specific clause that causes the deviation in combination with the deviation type.

[0152] Step S153: Modify the corresponding clause in the first reverse traceability rule, adjust the attribute feature matching criteria for the reverse mapping of attribute features from the recycling stage to the usage stage, and the adjusted attribute feature matching criteria conform to the actual attribute feature transfer situation reflected in the evaluation deviation trajectory.

[0153] The rule dynamic adjustment module is activated, loading candidate problem clauses and evaluating the deviation trajectory and deviation value sequence. A closed-loop control algorithm is used to adjust the matching standard, taking the deviation value as input and the matching standard as output, and calculating the adjustment amount through a proportional-integral-derivative controller. When the deviation type is normal, fuzzy control compensation is enabled, dividing the adjustment amount into multiple fuzzy subsets, and calculating the weight of each subset through the deviation membership function. The adjusted clauses are then simulated and verified by the rule verification engine to ensure that the score improvement in conformity with the actual attribute feature transmission is ≥S% (S is the set improvement threshold).

[0154] Step S154: For the deviation at the second tracing node, analyze the degree of fit between the attribute feature matching standard and the actual attribute feature in the second reverse tracing rule, and find the rule clause that caused the deviation.

[0155] Using the same rule matching degree calculation module as in step S152, the second reverse tracing rule text and the actual attribute feature data stream of the second tracing node are input. When parsing the rules, the process conversion rule library is loaded as a reference, and dimensions such as process fluctuation coefficient and equipment state vector are added when extracting actual feature parameters. Clause positioning focuses on matching clauses related to key conversion attribute features. When the actual matching degree is consistently below the threshold and the deviation propagation direction meets specific conditions, it is marked as a problem clause.

[0156] Step S155: Modify the corresponding clause in the second reverse tracing rule, adjust the attribute feature matching standard for reverse mapping of stage attribute features to the production and manufacturing stage, and the adjusted attribute feature matching standard is consistent with the actual attribute feature transmission situation reflected in the evaluation deviation trajectory.

[0157] Based on the key transformation attribute characteristics of the second traceability node, a multi-objective optimization algorithm is used to adjust the matching criteria. The dual objective functions are minimizing the deviation value and minimizing the sensitivity to process parameters, with the constraint that the adjusted matching criteria do not cause the deviation value of other nodes to increase by more than T% (T being a set upper limit). The optimized matching criteria clauses are obtained through iterative solution using an evolutionary algorithm, and the rule base is updated after simulation verification.

[0158] Step S156: For the deviation at the third tracing node, analyze the degree of fit between the attribute feature matching standard and the actual attribute feature in the third reverse tracing rule, and find the rule clause that caused the deviation.

[0159] Call the rule matching degree calculation module, and input the third reverse traceability rule text and the actual attribute feature data stream of the third traceability node. When parsing the rule, associate with the mineral resource characteristic database, and extract the actual characteristic parameters including the fluctuation coefficient of ore grade, the stability index of ore dressing recovery rate, etc. The clause positioning focuses on checking the scaling clause related to the raw material purity reference value. When the actual fluctuation range exceeds the threshold and shows a seasonal change trend, it is marked as a problematic clause.

[0160] Step S157: Modify the corresponding clause in the third reverse traceability rule, adjust the attribute feature matching standard for the reverse mapping of the attribute features in the production and manufacturing stage to the raw material acquisition stage. The adjusted attribute feature matching standard conforms to the actual attribute feature transfer situation reflected in the evaluation deviation trajectory.

[0161] Based on the raw material purity reference value feature of the third traceability node, use the time series prediction algorithm to predict the future attribute feature fluctuation trend. Take the prediction result as the reference for adjusting the matching standard, set the dynamic threshold by time period, and introduce the raw material quality grade correction factor. When the raw material quality reaches the specified grade, the requirement for the fluctuation range can be relaxed by a certain proportion. Compare and verify the adjusted clause with the historical data of the mine resource management system to ensure that it covers most of the raw material fluctuation conditions.

[0162] Step S158: Analyze the node attribute feature mismatch information, and determine the attribute feature response thresholds that mismatch with the core attribute features of the traceability node in the usage stage evaluation link, the production and manufacturing stage evaluation link, and the raw material acquisition stage evaluation link.

[0163] Start the threshold sensitivity analysis module, and load the node feature mismatch information and the configuration files of each stage of the evaluation model. Use the grey relational analysis method to calculate the correlation degree between the feature response thresholds of each evaluation link and the core attribute features of the node. The thresholds with a correlation degree > U (U is the set correlation degree threshold, 0 < U < 1) are included in the candidate set. Calculate the mismatch contribution degree for the thresholds in the candidate set: mismatch contribution degree = (number of mismatch feature dimensions / total number of feature dimensions) × (mismatch degree / maximum mismatch degree) × V% (V is a constant). When the mismatch contribution degree > W% (W is the set contribution degree threshold), it is determined as the sensitive threshold causing the mismatch.

[0164] Step S159: Adjust the attribute feature response threshold in the usage stage evaluation link, and the adjusted attribute feature response threshold covers the effective data range of the core attribute features of the first traceability node; adjust the attribute feature response threshold in the production and manufacturing stage evaluation link, and the adjusted attribute feature response threshold covers the effective data range of the core attribute features of the second traceability node; adjust the attribute feature response threshold in the raw material acquisition stage evaluation link, and the adjusted attribute feature response threshold covers the effective data range of the core attribute features of the third traceability node.

[0165] A threshold adjustment decision matrix is ​​constructed, where the row vectors represent the threshold IDs to be adjusted, and the column vectors represent the adjustment constraints (effective data range coverage, evaluation accuracy loss rate, and computational complexity change rate). The weights of each constraint are determined using the analytic hierarchy process (AHP). Multiple candidate threshold adjustment schemes are generated through Monte Carlo simulation, and the comprehensive score of each scheme is calculated. The scheme with the highest score is selected for adjustment. The adjustment process uses gradient descent to gradually approach the target threshold. After each adjustment, performance metrics are calculated using the model validation set. When the performance metric drops below a set threshold, a fallback mechanism is triggered.

[0166] Step S1591: Extract the core attribute feature parameters of the first traceability node, analyze the distribution range, the changing trend, and the core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features.

[0167] Core parameters are extracted from the attribute feature data lake of the first source node using a feature parameter extractor. A parameter distribution curve is plotted using kernel density estimation, with a confidence level set at X% (X being the statistical confidence level). The area under the distribution curve at X% is defined as the distribution range. The slope of the linear regression equation is used to determine the trend: slope > 0 indicates a positive trend, slope < 0 indicates a negative trend, and slope = 0 indicates a stationary trend. An improved K-means clustering algorithm is used to determine the core value range. The number of clusters is optimized using the silhouette coefficient method. The core value range is defined as the range containing the cluster centers and including Y% (Y being a set proportion) of data. The effective data range = distribution range ∩ (core value range ± Z times the standard deviation of the core value range) (Z being the safety factor).

[0168] Step S1592: Obtain the current attribute feature response threshold in the evaluation phase. The value standard of the attribute feature response threshold and the applicable attribute feature parameter range are extracted and presented through the original configuration file of the evaluation model.

[0169] The configuration file parser reads the evaluation model configuration file, locates the threshold label corresponding to the evaluation stage, extracts the specific value of the threshold, the value standard (such as based on experience value, historical data statistics) and the applicable parameter range, and converts it into a structured data format. The value standard includes subfields such as statistical basis, confidence level, and update cycle, and the applicable parameter range is presented in the form of interval.

[0170] Step S1593: Compare the effective data range of the core attribute features with the applicable range of the attribute feature response threshold, identify the overlapping and difference areas between the two, and analyze the impact of the difference areas on the evaluation results.

[0171] Compare the effective data range of the core attribute features of the first traceability node with the applicable range of the current feature response threshold in the evaluation stage, and draw a range distribution map to identify overlapping areas (areas covered by both effective data and the applicable threshold range) and discrepancy areas (areas covered only by effective data or only by the applicable threshold range). Analyze the impact of discrepancy areas on the evaluation results, such as whether discrepancy areas lead to valid data being misjudged as invalid or invalid data being misjudged as valid. The degree of impact is represented by the error propagation coefficient, which is calculated as: coefficient = (data volume in the discrepancy area / total data volume) × average relative error.

[0172] Step S1594: Based on the distribution of attribute feature parameters in the difference region, deduce the direction and range of the attribute feature response threshold that needs to be adjusted, so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features.

[0173] The characteristic parameters of the difference region are fitted with a probability distribution. The optimal distribution model is determined through a goodness-of-fit test, and the adjustment margin is calculated. The adjustment amount for the left-side difference region is calculated as: (lower limit of the effective data range - lower limit of the threshold's applicable range). When the lower limit of the effective data range is less than the lower limit of the threshold's applicable range, the lower limit of the threshold needs to be lowered by the adjustment amount × (1 + safety factor). The adjustment amount for the right-side difference region is calculated as: (upper limit of the threshold's applicable range - upper limit of the effective data range). When the upper limit of the threshold's applicable range is greater than the upper limit of the effective data range, the upper limit of the threshold needs to be increased by the adjustment amount × (1 + safety factor). The applicable range of the adjusted threshold is solved using an optimization algorithm. The objective function is to minimize the sum of squared errors in the adjusted threshold evaluation, and the constraint is that the adjusted range is ≤ a certain proportion of the original threshold range.

[0174] Step S1595: According to the derived adjustment direction and range, modify the attribute characteristic response threshold parameters of the usage stage evaluation link, and record the values ​​of the attribute characteristic response threshold before and after the adjustment and the basis for the adjustment.

[0175] The system connects to the threshold management database via the evaluation model parameter configuration interface, initiates a transaction processing mechanism, and executes update statements to modify threshold parameters, ensuring concurrency safety. Adjustments are recorded in a structured log format, including fields such as threshold values ​​before and after adjustment, difference region analysis report ID, fitted distribution model parameters, safety factor calculation process, and optimization algorithm iteration records. The log is stored on a blockchain, with the block hash value written to the header of the evaluation model configuration file. Upon completion of adjustments, a threshold synchronization mechanism is triggered, broadcasting the update event to all computing nodes. Nodes then reload the threshold parameters and execute self-check procedures.

[0176] Step S1596: Extract the core attribute feature parameters of the second traceability node, analyze the distribution range, the changing trend, and the core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features.

[0177] Using the same feature parameter extraction process as step S1591, for the process parameter-performance correlation features of the second traceability node, a process stability weighting factor (weight = 1 / process fluctuation coefficient) is added during kernel density estimation. The core value range is determined by introducing a process capability index, and the core value range is adjusted according to the magnitude of the process capability index. Data points from periods of process anomalies are excluded from the effective data range.

[0178] Step S1597: Obtain the current attribute characteristic response threshold in the evaluation stage of the production and manufacturing phase. The value standard of the attribute characteristic response threshold and the applicable attribute characteristic parameter range are extracted and presented through the original configuration file of the evaluation model.

[0179] Read the threshold configuration of the production and manufacturing stage evaluation link in the configuration file, and extract the threshold value, value standard, and applicable parameter range. The applicable parameter range includes the equipment status code subfield. Extract the applicable parameter range for the corresponding state according to the equipment operating status. The value standard includes statistical basis such as process verification data.

[0180] Step S1598: Compare the effective data range of the core attribute features with the applicable range of the attribute feature response threshold to find the overlapping and difference areas between the two.

[0181] The comparison method is the same as step S1593. For the continuous process characteristics of the production and manufacturing stage, a sliding time window interval comparison is adopted. The window size is the process batch cycle. The difference area analysis adds a equipment utilization rate correction factor and adjusts the weight of the difference area according to the utilization rate.

[0182] Step S1599: Based on the distribution of attribute feature parameters in the difference region, deduce the direction and range of the attribute feature response threshold that needs to be adjusted so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features.

[0183] The derivation process introduces a process capability index correction term, with the adjustment amount = basic adjustment amount × (1 + (standard process capability index - actual process capability index) / standard process capability index). The adjustment range constraint is increased to "threshold fluctuation range ≤ process parameter fluctuation range".

[0184] Step S15910: Modify the attribute characteristic response threshold parameters of the evaluation stage in the production and manufacturing phase according to the derived adjustment direction and range, and record the attribute characteristic response threshold adjustment information.

[0185] The modification process is similar to step S1595. The transaction processing synchronously locks the corresponding process parameter database table. The adjustment basis includes comparative analysis of process verification reports. The log records add fields for equipment number and process batch number.

[0186] Step S15911: Extract the core attribute feature parameters of the third source node, analyze the distribution range, changing trend and core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features.

[0187] In response to the characteristics of mineral resources in the raw material acquisition stage, the parameter extraction adds a grade correction coefficient (actual grade / standard grade), the distribution range calculation adopts robust estimation to resist the influence of outliers, and the core value range is combined with the definition of mine recoverable reserves to ensure that it includes the attribute characteristic parameters corresponding to most of the recoverable reserves. The effective data range is filtered out by resource and environmental carrying capacity assessment to remove parameter points that exceed the carrying capacity threshold.

[0188] Step S15912: Obtain the current attribute response threshold of the raw material acquisition stage evaluation process. The value standard of the attribute response threshold and the applicable attribute parameter range are extracted and presented through the original configuration file of the evaluation model.

[0189] When reading the configuration file, the mining area ID and resource type fields are extracted. The threshold configurations for different mining areas and resource types are stored independently, and the value standards include statistical basis such as mine production plan data.

[0190] Step S15913: Compare the effective data range of the core attribute features with the applicable range of the attribute feature response threshold to identify the overlapping and difference areas between the two.

[0191] When making comparisons, a resource recovery rate weight is introduced, and the criteria for determining overlapping areas are adjusted based on the resource recovery rate. The analysis of differences in areas is combined with the stages of the mine's life cycle, and different weights are assigned to different areas at different stages.

[0192] Step S15914: Based on the distribution of attribute feature parameters in the difference region, deduce the direction and range of the attribute feature response threshold that needs to be adjusted, so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features.

[0193] The derivation process incorporates a resource price volatility coefficient as an adjustment factor, and the adjustment range is adjusted based on the resource price volatility, with the adjustment direction conforming to the trend of the mine production plan.

[0194] Step S15915: Modify the attribute characteristic response threshold parameters of the raw material acquisition stage evaluation link according to the derived adjustment direction and range, and record the attribute characteristic response threshold adjustment information.

[0195] The modification process follows step S1595, and the adjustment basis includes mining resource planning data. The log records are updated with mining area ID and resource type fields to ensure they are associated with specific mineral resource characteristics.

[0196] Step S1510: Re-embed the adjusted first reverse tracing rule, second reverse tracing rule, and third reverse tracing rule into the green attribute reverse calibration component set, and synchronize the adjusted attribute characteristic response thresholds of each stage of evaluation to the evaluation model to form an optimized green product evaluation model.

[0197] The adjusted reverse tracing rules were recompiled into a rule configuration file and embedded into a set of green attribute reverse calibration components. The correctness of the rule logic was verified through component interface testing. The adjusted attribute characteristic response thresholds for each stage of the evaluation were synchronized to the corresponding modules of the model through the evaluation model parameter configuration interface. Model integration testing was performed to verify the improved accuracy of the evaluation results, ultimately forming an optimized green product evaluation model.

[0198] Figure 2 The illustration shows the hardware structure of a big data-based green product evaluation model optimization system 100 provided in an embodiment of the present invention for implementing the above-described big data-based green product evaluation model optimization method. Figure 2 As shown, the green product evaluation model optimization system 100 based on big data may include a processor 110, a machine-readable storage medium 120, a bus 130, and a communication unit 140.

[0199] Machine-readable storage medium 120 can store data and / or instructions. In a specific implementation, one or more processors 110 execute the computer-executable instructions stored in the machine-readable storage medium 120, enabling the processor 110 to execute the big data-based green product evaluation model optimization method as described in the above method embodiment. The processor 110, machine-readable storage medium 120, and communication unit 140 are connected via bus 130, and the processor 110 can be used to control the transmission and reception actions of the communication unit 140. The specific implementation process of the processor 110 can be found in the various method embodiments executed by the big data-based green product evaluation model optimization system 100 described above, and their implementation principles and technical effects are similar, so they will not be repeated here.

[0200] Furthermore, this embodiment of the invention also provides a readable storage medium containing computer-executable instructions. When the processor executes the computer-executable instructions, the above-mentioned optimization method for the green product evaluation model based on big data is implemented.

[0201] It should be noted that, in order to simplify the description of this invention and thus aid in the understanding of one or more embodiments, the foregoing description of the embodiments of this invention sometimes combines multiple features into a single embodiment, drawing, or description thereof. Similarly, it should be noted that, in order to simplify the description of this invention and thus aid in the understanding of one or more embodiments, the foregoing description of the embodiments of this invention sometimes combines multiple features into a single embodiment, drawing, or description thereof.

Claims

1. A method for optimizing a green product evaluation model based on big data, characterized in that, The method includes: Collect reverse traceability green attribute data for the entire life cycle of green products. The reverse traceability green attribute data includes data on the purity of recycled resources, energy consumption control data, actual energy consumption data, loss and degradation data, actual emission reduction data of the manufacturing process, raw material conversion efficiency data, environmental impact data of mining, and resource utilization rate data during the raw material acquisition stage. The reverse mining process is used to identify the attribute relationships and traceability nodes between the recycling stage and the usage stage, the usage stage and the production and manufacturing stage, and the production and manufacturing stage and the raw material acquisition stage in the reverse traceability green attribute data. The traceability nodes are the key feature intersection points in the reverse attribute transmission process. Based on the attribute association and tracing nodes, a set of green attribute reverse calibration components for the evaluation model is constructed. The set of green attribute reverse calibration components includes reverse tracing rules, node feature mapping relationships, and stage attribute feedback channels. The reverse traceability green attribute data is input into the green attribute reverse calibration component set according to the traceability node, and is transmitted back to the previous stage evaluation link of the evaluation model through the stage attribute feedback channel to generate green product evaluation deviation trajectory and node feature mismatch information. Based on the evaluation deviation trajectory and node feature mismatch information, the reverse tracing rules in the green attribute reverse calibration component set and the feature response thresholds of each stage of the evaluation model are dynamically adjusted to obtain the optimized green product evaluation model.

2. The method for optimizing a green product evaluation model based on big data according to claim 1, characterized in that, The reverse mining of the attribute associations and traceability nodes in the reverse traceability green attribute data between the recycling stage and the usage stage, the usage stage and the production and manufacturing stage, and the production and manufacturing stage and the raw material acquisition stage includes: Using the purity data of recycled resources and the energy consumption control data of recycled resources in the recycling stage as the starting data source, the attribute features related to the use stage are extracted. The extracted attribute features related to the use stage reflect the impact of the product status in the use stage on the product attributes in the recycling stage. Track the transmission trajectory of extracted attribute features related to the usage stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the actual energy consumption data and loss degradation data of the usage stage, and establish a preliminary attribute association between the recycling stage and the usage stage. Analyze the interaction of attribute features in the preliminary attribute association, identify the core attribute features that remain stable during the transmission of attribute features, and determine the intersection of the core attribute features as the first traceability node between the recycling stage and the usage stage; Using actual energy consumption data and loss / degradation data during the usage phase as intermediate data sources, attribute features related to the manufacturing phase are extracted. The extracted attribute features related to the manufacturing phase reflect the impact of process parameters during the manufacturing phase on product attributes during the usage phase. Track the transmission trajectory of extracted attribute features related to the production and manufacturing stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the actual emission reduction data and raw material conversion efficiency data of the production and manufacturing stage, and establish the attribute association between the use stage and the production and manufacturing stage. Analyze the transmission logic of attribute feature items in the attribute association, identify the key transformation attribute features when attribute feature items are transferred from the manufacturing stage to the use stage, and determine the intersection of the key transformation attribute features as the second traceability node between the use stage and the manufacturing stage. Using actual emission reduction data and raw material conversion efficiency data from the production and manufacturing stage as intermediate data sources, attribute features related to the raw material acquisition stage are extracted. The extracted attribute features related to the raw material acquisition stage reflect the impact of resource data from the raw material acquisition stage on material properties in the production and manufacturing stage. Track the transmission trajectory of extracted attribute features related to the raw material acquisition stage in the reverse traceability green attribute data, locate the corresponding attribute feature items in the mining environmental impact data and resource utilization rate data of the raw material acquisition stage, and establish the attribute association between the production and manufacturing stage and the raw material acquisition stage. Analyze the transmission basis of attribute feature items in this attribute association, identify the core supporting attribute features when attribute feature items are transmitted from the raw material acquisition stage to the production and manufacturing stage, and determine the intersection of the core supporting attribute features as the third traceability node between the production and manufacturing stage and the raw material acquisition stage. The attribute relationships between the recycling stage and the usage stage, the attribute relationships between the usage stage and the production and manufacturing stage, the attribute relationships between the production and manufacturing stage and the raw material acquisition stage, as well as the first traceability node, the second traceability node, and the third traceability node, are integrated to form a reverse attribute relationship and a set of traceability nodes.

3. The method for optimizing a green product evaluation model based on big data according to claim 1, characterized in that, The set of green attribute reverse calibration components for constructing the evaluation model based on the attribute association and source node includes: Analyze the attribute features of the first, second, and third source nodes, and extract and present the core attribute feature type, core attribute feature transmission direction, and core attribute feature interaction method for each source node; To address the reverse propagation requirement from the recycling phase to the usage phase, a first reverse tracing rule is generated. The content of the first reverse tracing rule includes the method of reverse mapping of attribute features from the recycling phase to the usage phase, the attribute feature matching criteria, and the attribute feature propagation priority. To address the reverse transmission requirement from the usage stage to the manufacturing stage, a second reverse traceability rule is generated. The content of the second reverse traceability rule includes the method of reverse mapping of attribute features from the usage stage to the manufacturing stage, the attribute feature matching criteria, and the attribute feature transmission priority. To address the reverse transmission requirement from the manufacturing stage to the raw material acquisition stage, a third reverse traceability rule is generated. The content of the third reverse traceability rule includes the method of reverse mapping of the attribute characteristics of the manufacturing stage to the raw material acquisition stage, the attribute characteristic matching standard, and the attribute characteristic transmission priority. Establish a node attribute feature mapping relationship, extract and present the correspondence between the core attribute features of the first traceability node and the attribute features of the usage stage evaluation link, the correspondence between the core attribute features of the second traceability node and the attribute features of the production and manufacturing stage evaluation link, and the correspondence between the core attribute features of the third traceability node and the attribute features of the raw material acquisition stage evaluation link. Plan and construct a first-stage attribute feedback channel, which connects the attribute data storage area of ​​the recycling stage with the evaluation stage of the usage stage. The first-stage attribute feedback channel is equipped with attribute feature transmission buffer processing, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the first reverse tracing rule. Plan and construct a second-stage attribute feedback channel, which connects the usage stage attribute data storage area with the production and manufacturing stage evaluation link. The second-stage attribute feedback channel is equipped with attribute feature transmission buffer processing, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the second reverse tracing rule. Plan and construct a third-stage attribute feedback channel, which connects the attribute data storage area of ​​the production and manufacturing stage with the evaluation link of the raw material acquisition stage. The attribute feature transmission buffer processing is set up in the third-stage attribute feedback channel, and the data format and transmission rate of the buffer processing are consistent with the data processing requirements defined by the third reverse traceability rule. A reverse calibration attribute feature integration process is established, which summarizes the reverse attribute feature data transmitted from the attribute feedback channels in three stages, and sorts and integrates the attribute features according to the order of the traceability nodes. The first reverse tracing rule, the second reverse tracing rule, the third reverse tracing rule, the node attribute feature mapping relationship, the first stage attribute feedback channel, the second stage attribute feedback channel, the third stage attribute feedback channel, and the reverse calibration attribute feature integration process are systematically integrated to form a green attribute reverse calibration component set.

4. The method for optimizing a green product evaluation model based on big data according to claim 1, characterized in that, Based on the evaluation deviation trajectory and node feature mismatch information, the reverse tracing rules in the green attribute reverse calibration component set and the feature response thresholds of each stage of the evaluation model are dynamically adjusted to obtain the optimized green product evaluation model, including: Analyze and evaluate the deviation trajectory, locate the distribution of deviation at the first, second, and third traceability nodes, and extract and present the deviation type and deviation transmission direction at each traceability node; To address the deviation at the first tracing node, analyze the degree of fit between the attribute feature matching standard and the actual attribute features in the first reverse tracing rule, and identify the rule clauses that caused the deviation. Modify the corresponding clauses in the first reverse tracing rule, adjust the attribute feature matching standard for the reverse mapping of attribute features from the recycling stage to the usage stage, and the adjusted attribute feature matching standard is consistent with the actual attribute feature transmission situation reflected in the evaluation deviation trajectory. To address the deviation at the second tracing node, analyze the degree of fit between the attribute feature matching standard and the actual attribute features in the second reverse tracing rule, and identify the rule clauses that caused the deviation. Modify the corresponding clauses in the second reverse tracing rule, and adjust the attribute feature matching standard for reverse mapping of stage attribute features to the production and manufacturing stage. The adjusted attribute feature matching standard is consistent with the actual attribute feature transmission situation reflected in the evaluation deviation trajectory. To address the deviation at the third tracing node, analyze the degree of fit between the attribute feature matching standard in the third reverse tracing rule and the actual attribute features, and identify the rule clauses that cause the deviation. Modify the corresponding clauses in the third reverse traceability rule, and adjust the attribute feature matching standard for the reverse mapping of attribute features from the production and manufacturing stage to the raw material acquisition stage. The adjusted attribute feature matching standard is consistent with the actual attribute feature transmission situation reflected in the evaluation deviation trajectory. Analyze the node attribute feature mismatch information to determine the attribute feature response threshold for mismatch with the core attribute features of the traceability node in the evaluation stages of the usage stage, the evaluation stage of the production and manufacturing stage, and the evaluation stage of the raw material acquisition stage; Adjust the attribute characteristic response thresholds in the usage stage evaluation process, so that the adjusted attribute characteristic response thresholds cover the effective data range of the core attribute characteristics of the first traceability node; adjust the attribute characteristic response thresholds in the production and manufacturing stage evaluation process, so that the adjusted attribute characteristic response thresholds cover the effective data range of the core attribute characteristics of the second traceability node; adjust the attribute characteristic response thresholds in the raw material acquisition stage evaluation process, so that the adjusted attribute characteristic response thresholds cover the effective data range of the core attribute characteristics of the third traceability node. The adjusted first, second, and third reverse traceability rules are re-embedded into the green attribute reverse calibration component set, and the adjusted attribute characteristic response thresholds of each stage of the evaluation are synchronized to the evaluation model to form an optimized green product evaluation model.

5. The method for optimizing a green product evaluation model based on big data according to claim 2, characterized in that, The process of tracing the transmission trajectory of the extracted attribute features related to the usage phase in the reverse-source green attribute data, and locating the corresponding attribute feature item in the actual energy consumption data and degradation data during the usage phase, includes: The extracted attribute features related to the usage stage are labeled with attribute features. The labeling content includes the attribute type of the attribute feature, the data source of the attribute feature, and the core expression form of the attribute feature, so that the attribute features are identifiable in the process of tracking the transmission trajectory. Starting from the attribute data storage area of ​​the recycling stage in reverse chronological order, retrieve the attribute feature data stream with the attribute feature mark in the reverse tracing green attribute data, and record the storage location and associated data identifier of the attribute feature data stream; Following the retrieved attribute feature data stream, the movement path of the attribute feature data stream in the reverse source green attribute data is tracked, and the intermediate storage units and data processing nodes passed through by the attribute feature data stream are recorded to form a preliminary transmission trajectory. In the initial transmission trajectory, the attribute feature data stream branches pointing to the data storage area of ​​the usage stage are selected, the transmission details of the attribute feature data stream branches are focused, and the entry position of the attribute feature data stream into the data storage area of ​​the usage stage is recorded. Within the data storage area during the usage phase, attribute features in the actual energy consumption data are retrieved by attribute feature marking. The attribute features in the retrieved actual energy consumption data are compared with the attribute types and core manifestations of the marked attribute features to identify similar attribute feature items. Using the same method, retrieve the attribute features of the loss and degradation data in the data storage area during the usage phase, compare the attribute features of the retrieved loss and degradation data with the attribute types and core manifestations of the marked attribute features, and find similar attribute feature items. Analyze the degree of correlation between similar attribute features and labeled attribute features. By analyzing the continuity of attribute feature transmission paths and the correlation of attribute feature parameters, determine the attribute features in the actual energy consumption data and the attribute features in the loss and degradation data that directly correspond to the labeled attribute features. Record the storage address, associated data group, and data generation time of the corresponding attribute feature item in the data storage area during the usage phase to form attribute feature item location information; By associating and integrating attribute feature labels, attribute feature transmission trajectories, and attribute feature item location information, the complete correspondence between the attribute features extracted during the recycling stage and those related to the usage stage and the attribute features in the actual energy consumption data and loss degradation data during the usage stage is extracted and presented. Based on the complete correspondence, the details of the attribute feature transmission trajectory are supplemented and improved, and the data form changes of attribute feature items during the transmission process are marked.

6. The method for optimizing a green product evaluation model based on big data according to claim 3, characterized in that, The establishment of node attribute feature mapping relationships, extracting and presenting the correspondence between the core attribute features of the first traceability node and the attribute features of the usage stage evaluation stage, the correspondence between the core attribute features of the second traceability node and the attribute features of the production and manufacturing stage evaluation stage, and the correspondence between the core attribute features of the third traceability node and the attribute features of the raw material acquisition stage evaluation stage, includes: Extract the core attribute features of the first traceability node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the first traceability node. All evaluation attribute features of the usage phase evaluation stage are obtained. The evaluation objectives, data requirements and parameter ranges of each evaluation attribute feature are extracted and presented from the design documents and data records of the corresponding evaluation stage, forming a list of evaluation attribute features for the usage phase. By comparing the core attribute feature list of the first traceability node with the attribute feature list of the usage phase evaluation, attribute feature combinations with consistent attribute parameter types and matching data types are identified to form preliminary corresponding candidate pairs; The functional correlation between the core attribute features of the first traceability node and the evaluation attribute features of the usage stage in the preliminary corresponding candidate pairs is analyzed. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the first traceability node and the data change trend of the evaluation attribute features of the usage stage. Valid corresponding candidate pairs are screened based on this correlation. Determine the mapping method for valid candidate pairs, extract and present how the core attribute features of the first traceability node are transformed into calibration reference data for evaluating attribute features in the usage phase, and form the first set of attribute feature correspondences; Extract the core attribute features of the second traceability node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the second traceability node. All evaluation attribute characteristics of the production and manufacturing stage evaluation process are obtained. The evaluation objectives, data requirements and parameter ranges of each evaluation attribute characteristic are extracted and presented through the design documents and data records of the corresponding evaluation stage, forming a list of evaluation attribute characteristics for the production and manufacturing stage. By comparing the core attribute feature list of the second traceability node with the evaluation attribute feature list of the manufacturing stage, attribute feature combinations with consistent attribute parameter types and matching data types are identified to form preliminary corresponding candidate pairs. The functional correlation between the core attribute features of the second traceability node and the evaluation attribute features of the production and manufacturing stage in the preliminary candidate pairs is analyzed. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the second traceability node and the data change trend of the evaluation attribute features of the production and manufacturing stage. Based on this correlation, valid candidate pairs are screened, the mapping method is determined, and a second set of attribute feature correspondences is formed. Extract the core attribute features of the third source node, extract and present the attribute parameters, the representation form and data type of the core attribute features, and form a list of core attribute features of the third source node. All assessment attribute characteristics of the raw material acquisition stage assessment process are obtained. The assessment objectives, data requirements, and parameter ranges of each assessment attribute characteristic are extracted and presented through the design documents and data records of the corresponding assessment process, forming a list of assessment attribute characteristics for the raw material acquisition stage. By comparing the core attribute feature list of the third traceability node with the evaluation attribute feature list of the raw material acquisition stage, attribute feature combinations with consistent attribute parameter types and matching data types are identified to form preliminary corresponding candidate pairs. The functional correlation between the core attribute features of the third traceability node and the evaluation attribute features of the raw material acquisition stage in the preliminary candidate pairs is analyzed. The functional correlation refers to the negative or positive correlation between the data change trend of the core attribute features of the third traceability node and the data change trend of the evaluation attribute features of the raw material acquisition stage. Based on this correlation, valid candidate pairs are screened, the mapping method is determined, and a third set of attribute feature correspondence is formed. Integrate the first set of attribute feature correspondences, the second set of attribute feature correspondences, and the third set of attribute feature correspondences to form a node attribute feature mapping relationship.

7. The method for optimizing a green product evaluation model based on big data according to claim 4, characterized in that, The adjustment of the attribute characteristic response thresholds in the usage stage evaluation process to match the core attribute characteristics of the first traceability node; the adjustment of the attribute characteristic response thresholds in the production and manufacturing stage evaluation process to match the core attribute characteristics of the second traceability node; and the adjustment of the attribute characteristic response thresholds in the raw material acquisition stage evaluation process to match the core attribute characteristics of the third traceability node include: Extract the core attribute feature parameters of the first traceability node, analyze the distribution range, changing trend and core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features. The current attribute feature response thresholds in the usage phase evaluation stage are obtained. The criteria for the attribute feature response thresholds and the applicable attribute feature parameter ranges are extracted and presented through the original configuration file of the evaluation model. By comparing the effective data range of the core attribute features with the applicable range of the attribute feature response threshold, we can identify the overlapping and difference areas between the two and analyze the impact of the difference areas on the evaluation results. Based on the distribution of attribute feature parameters in the difference region, the direction and range of attribute feature response threshold adjustment are derived so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features. Based on the derived adjustment direction and scope, modify the attribute characteristic response threshold parameters in the usage phase evaluation stage, and record the values ​​of the attribute characteristic response thresholds before and after the adjustment and the basis for the adjustment. Extract the core attribute feature parameters of the second traceability node, analyze the distribution range, changing trend and core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features. The current attribute response thresholds in the evaluation stage of the production and manufacturing process are obtained. The criteria for the attribute response thresholds and the applicable range of attribute parameters are extracted and presented through the original configuration file of the evaluation model. By comparing the effective data range of the core attribute features with the applicable range of the attribute feature response threshold, the overlapping and difference areas between the two can be identified. Based on the distribution of attribute feature parameters in the difference region, the direction and range of attribute feature response threshold adjustment are derived so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features. Based on the derived adjustment direction and scope, modify the attribute characteristic response threshold parameters of the evaluation stage in the production and manufacturing phase, and record the attribute characteristic response threshold adjustment information. Extract the core attribute feature parameters of the third traceability node, analyze the distribution range, changing trend and core value range of the core attribute feature parameters, and determine the effective data range of the core attribute features. The current attribute response thresholds in the raw material acquisition stage evaluation are obtained. The criteria for the attribute response thresholds and the applicable attribute parameter ranges are extracted and presented through the original configuration file of the evaluation model. By comparing the effective data range of the core attribute features with the applicable range of the attribute feature response threshold, the overlapping and difference areas between the two can be identified. Based on the distribution of attribute feature parameters in the difference region, the direction and range of attribute feature response threshold adjustment are derived so that the applicable range of the adjusted attribute feature response threshold can fully cover the effective data range of the core attribute features. Based on the derived adjustment direction and scope, modify the attribute characteristic response threshold parameters of the raw material acquisition stage evaluation process, and record the attribute characteristic response threshold adjustment information.

8. The method for optimizing a green product evaluation model based on big data according to claim 2, characterized in that, The analysis of the attribute feature item transmission logic in the attribute association identifies the key transformation attribute features when attribute feature items are transferred from the manufacturing stage to the usage stage, and determines the intersection of the key transformation attribute features as the second traceability node between the usage stage and the manufacturing stage, including: Deconstruct all attribute features in the attribute association between the usage stage and the manufacturing stage, classify them according to the direction of attribute feature transmission, and distinguish between attribute features output from the manufacturing stage and attribute features input to the usage stage. By comparing the attribute features output in the manufacturing stage with the attribute features input in the usage stage, the inherent logic of the attribute parameter changes is analyzed. The inherent logic describes the transformation process of attribute features from the manufacturing stage to the usage stage based on predefined process transformation rules and product form mapping relationships. Track the transformation process of output attribute features at each stage of production and manufacturing, record key operations such as attribute parameter adjustment, attribute feature addition, and attribute feature merging during the transformation process, and construct a transformation process record; In the conversion process record, identify the conversion operation that plays a decisive role in the formation of the attribute characteristics of the usage stage. The attribute characteristic change corresponding to the conversion operation is the key conversion attribute characteristic. Determine the location of key conversion attribute features in the conversion process record, and record the attribute feature data status, conversion operation type, and associated attribute feature items corresponding to that location; Analyze the relationship between key conversion attributes and other conversion attributes, and extract and present how key conversion attributes affect the conversion direction and degree of other attributes; Locate the storage location of key transformation attribute features in the reverse traceability green attribute data, and find the related attribute feature data around that location. The related attribute feature data includes related production and manufacturing stage attribute feature data and usage stage attribute feature data. Determine the intersection region of the associated attribute feature data, which contains the complete transformation trajectory of the key transformation attribute features and the interaction information of the associated attribute feature items; The intersection area is defined as the second traceability node between the usage stage and the manufacturing stage. The interactive data of the coverage of the second traceability node includes key transformation attribute features and related attribute features. Record the node identifier, coverage, core key transformation attributes, and related attribute features of the second traceability node.

9. The method for optimizing a green product evaluation model based on big data according to claim 3, characterized in that, The first-stage attribute feedback channel is planned and constructed. This channel connects the attribute data storage area of ​​the recycling stage with the evaluation stage of the usage stage. The attribute feedback channel includes an attribute feature transfer buffer to adapt to the execution of the first reverse tracing rule, including: Plan the physical data transmission path between the attribute data storage area in the recycling phase and the evaluation stage in the usage phase. The path design should avoid unnecessary data processing nodes and shorten the transmission distance. A data receiving port is set in the transmission path. This data receiving port is matched with the output interface of the attribute data storage area in the recycling stage and is able to receive attribute feature data that meets the requirements of the first reverse tracing rule. An attribute feature filtering unit is set up after the data receiving port. This attribute feature filtering unit performs preliminary filtering on the received attribute feature data according to the attribute feature matching standard in the first reverse tracing rule, and retains the attribute feature data that meets the requirements of the first reverse tracing rule. An attribute feature sorting unit is set after the attribute feature filtering unit. The attribute feature sorting unit arranges the filtered attribute feature data in order according to the attribute feature transmission priority in the first reverse tracing rule, forming an ordered attribute feature data stream. A buffer storage unit is set after the attribute feature sorting unit. This buffer storage unit provides temporary storage space for the ordered attribute feature data stream, reducing the loss of attribute feature data caused by data transmission congestion. A data preservation method is set in the buffer storage unit to maintain the original attribute parameters of the attribute feature data without changing the core attribute features and relationships of the attribute feature data. A data sending unit is set up after the buffer storage unit. This data sending unit sends ordered attribute feature data streams to the use stage evaluation stage in batches according to the processing rhythm of the use stage evaluation stage. A status synchronization unit is set at both ends of the attribute feedback channel to synchronize the output status of the attribute data storage area in the recycling phase with the receiving status in the evaluation phase in real time, thereby achieving data transmission synchronization. An attribute feature verification process is set up inside the attribute feedback channel. The integrity of attribute feature data during transmission is presented by comparing attribute feature parameters, thereby reducing transmission errors. The data receiving port, attribute feature filtering unit, attribute feature sorting unit, buffer storage unit, data preservation processing method, data sending unit, status synchronization unit, and attribute feature verification processing flow are integrated into the planned transmission path to form the first-stage attribute feedback channel.

10. A green product evaluation model optimization system based on big data, characterized in that, The big data-based green product evaluation model optimization system includes a processor and a memory, the memory and the processor being connected. The memory is used to store programs, instructions or code, and the processor is used to run the programs, instructions or code in the memory to implement the big data-based green product evaluation model optimization method according to any one of claims 1-9.

Images